Journal of

Personalized
Medicine

Review
Integrating Machine Learning with Multi-Omics Technologies in
Geroscience: Towards Personalized Medicine
Nikolaos Theodorakis 1,2 , Georgios Feretzakis 3 , Lazaros Tzelves 4 , Evgenia Paxinou 3 , Christos Hitas 1 ,
Georgia Vamvakou 1 , Vassilios S. Verykios 3, * and Maria Nikolaou 1

1 Department of Cardiology & 65+ Clinic, Amalia Fleming General Hospital, 14, 25th Martiou Str.,
15127 Melissia, Greece; [email protected] (N.T.); [email protected] (C.H.);
[email protected] (G.V.); [email protected] (M.N.)
2 School of Medicine, National and Kapodistrian University of Athens, 75 Mikras Asias, 11527 Athens, Greece
3 School of Science and Technology, Hellenic Open University, 18 Aristotelous Str., 26335 Patras, Greece;
[email protected] (G.F.); [email protected] (E.P.)
4 2nd Department of Urology, Sismanoglio General Hospital, Sismanogliou 37, National and Kapodistrian
University of Athens, 15126 Athens, Greece; [email protected]
* Correspondence: [email protected]; Tel.: +30-2610367698

Abstract: Aging is a fundamental biological process characterized by a progressive decline in physio-

logical functions and an increased susceptibility to diseases. Understanding aging at the molecular
level is crucial for developing interventions that could delay or reverse its effects. This review
explores the integration of machine learning (ML) with multi-omics technologies—including ge-
nomics, transcriptomics, epigenomics, proteomics, and metabolomics—in studying the molecular
hallmarks of aging to develop personalized medicine interventions. These hallmarks include genomic
instability, telomere attrition, epigenetic alterations, loss of proteostasis, disabled macroautophagy,
deregulated nutrient sensing, mitochondrial dysfunction, cellular senescence, stem cell exhaustion,
altered intercellular communication, chronic inflammation, and dysbiosis. Using ML to analyze big
and complex datasets helps uncover detailed molecular interactions and pathways that play a role in
aging. The advances of ML can facilitate the discovery of biomarkers and therapeutic targets, offering
Citation: Theodorakis, N.; Feretzakis, insights into personalized anti-aging strategies. With these developments, the future points toward a
G.; Tzelves, L.; Paxinou, E.; Hitas, C.; better understanding of the aging process, aiming ultimately to promote healthy aging and extend
Vamvakou, G.; Verykios, V.S.; life expectancy.
Nikolaou, M. Integrating Machine
Learning with Multi-Omics
Keywords: machine learning; multi-omics technologies; aging research; geroscience; hallmarks of
Technologies in Geroscience: Towards
aging; personalized medicine
Personalized Medicine. J. Pers. Med.
2024, 14, 931. https://doi.org/
10.3390/jpm14090931

Academic Editor: Moon-Soo Lee 1. Introduction

Received: 10 August 2024
1.1. Background on Aging and Its Medical Significance
Revised: 27 August 2024 Aging is an inevitable biological process marked by a progressive decline in cellular
Accepted: 30 August 2024 and physiological functions. This decline is associated with increased vulnerability to a
Published: 31 August 2024 range of chronic diseases, including cardiovascular diseases (CVDs), neurodegenerative
disorders, diabetes, and cancers [1]. Aging not only affects the health and quality of life of
individuals but also poses significant socioeconomic challenges due to rising healthcare
costs and the need for long-term care [2].
Copyright: © 2024 by the authors.
The study of aging, or geroscience, aims to understand the underlying mechanisms
Licensee MDPI, Basel, Switzerland.
that drive this process. The primary focus of geroscience is to identify strategies to mitigate
This article is an open access article
the complications of aging, thereby extending the period of healthy disease-free life, known
distributed under the terms and
as health span and overall lifespan [1]. This pursuit has led to the identification of a set of
conditions of the Creative Commons
Attribution (CC BY) license (https://
biological processes termed the hallmarks of aging. These hallmarks represent the common
creativecommons.org/licenses/by/
molecular mechanisms that mediate the aging process across different species and include
4.0/). genomic instability, telomere attrition, epigenetic alterations, loss of proteostasis, disabled

J. Pers. Med. 2024, 14, 931. https://doi.org/10.3390/jpm14090931 https://www.mdpi.com/journal/jpm

J. Pers. Med. 2024, 14, 931 2 of 27

macroautophagy, deregulated nutrient sensing, mitochondrial dysfunction, cellular senes-

cence, stem cell exhaustion, altered intercellular communication, chronic inflammation,
and dysbiosis [2].

1.2. Overview of Multi-Omics Data

The advances in omics technologies have revolutionized biological research, providing
comprehensive datasets that capture different layers of biological information. Multi-omics
approaches, which integrate data from various omics technologies, offer a holistic view of
the molecular mechanisms underlying complex biological processes like aging. Genomics
focuses on the study of the entire genome, including genes and regulatory elements. It
provides insights into genetic predispositions and mutations that contribute to aging and
age-related diseases [3]. Epigenomics examines modifications to the genetic material
that affect gene expression without altering the DNA sequence. These include DNA
methylation, histone modification, and non-coding RNA molecules. Epigenetic changes
are crucial in regulating gene expression and are significantly impacted by aging [4].
Transcriptomics involves the study of RNA transcripts produced by the genome. It reflects
gene expression and provides information on how genetic information is translated into
functional proteins. Transcriptomic analysis helps in understanding gene regulation and
its alterations during aging [5]. Proteomics is the large-scale study of proteins, including
their three-dimensional structure, function, and interactions. Proteins are the primary
effectors of cellular functions, and proteomic analyses reveal changes in protein expression,
post-translational modifications, and interactions that occur in the context of aging [6].
Metabolomics is the comprehensive analysis of metabolites, the small molecules involved
in metabolic pathways. Metabolomic profiling provides insights into the biochemical
activities and metabolic state of the cell and therefore organism, which can reveal the
metabolic shifts associated with aging [7].

1.3. Importance of Machine Learning in Modern Biomedical Research

Machine learning (ML) has become an indispensable tool in biomedical research,
particularly for handling and interpreting complex high-dimensional datasets generated
by omics technologies. The main strength of ML is its ability to identify patterns within
data and use them to make predictions or classifications.
ML algorithms are adept at managing large-scale and heterogeneous data, identifying
intricate patterns, and extracting meaningful insights. This capability is particularly useful
in integrating multi-omics data to understand the multifaceted nature of aging [8].
ML has been applied to various aspects of aging research, including the identification
of biomarkers, prediction of disease onset and progression, and the development of per-
sonalized treatment strategies. For instance, ML models can predict biological age based
on omics data, identify key molecular drivers of aging, and suggest potential therapeutic
targets [9].

1.4. Objectives
This review aims to provide a comprehensive overview of how integrating multi-
omics data with ML techniques can enhance our understanding of the hallmarks of aging.
We will discuss the role of different omics technologies in studying each hallmark of
aging. Furthermore, we will explore how ML models can be applied to predict and
analyze age-related changes. Additionally, we will highlight case studies and applications
that demonstrate the potential of this integrative approach. Moreover, we will report
the challenges and limitations in the field. Finally, we will suggest future directions for
advancing aging research and developing personalized strategies to extend health span
and lifespan.
J. Pers. Med. 2024, 14, 931 3 of 27

2. Hallmarks of Aging and Multi-Omics Data Integration

2.1. Genomic Instability
2.1.1. Role of Genetic Variations and Mutations in Aging
Genomic instability is a primary hallmark of aging, characterized by an increased
frequency of mutations and structural alterations in the DNA. These mutations arise from
DNA replication errors, environmental insults, and defective DNA repair mechanisms. Key
sources of genomic instability include the accumulation of reactive oxygen species (ROS),
which cause oxidative damage to DNA, and the decline in the efficiency of DNA repair path-
ways such as base excision repair (BER), nucleotide excision repair (NER), mismatch repair
(MMR), and double-strand break repair mechanisms including homologous recombination
and non-homologous end joining [10]. Studies have shown that aged cells exhibit higher
levels of DNA damage and chromosomal aberrations, which can drive oncogenesis and
cellular dysfunction. For instance, increased rates of somatic mutations in mitochondrial
DNA (mtDNA) have been linked to reduced cellular respiration and increased oxidative
stress, further exacerbating genomic instability [11]. Experimental evidence from model
organisms, such as mice with impaired DNA repair capabilities, demonstrates accelerated
aging phenotypes, underscoring the critical role of genomic integrity in longevity [12].

2.1.2. Machine Learning Models to Predict Genomic Instability and Its Effects
ML models are particularly useful in predicting the effects of genomic instability by
analyzing large-scale genomic datasets. Supervised learning ML algorithms, like support
vector machines (SVMs) and neural networks, can be trained on labeled genomic data
to classify mutations as benign or pathogenic, which is critical for understanding which
mutations contribute to aging and age-related diseases. Unsupervised learning techniques
like clustering can identify novel mutational signatures that may not be evident through
traditional analysis. These signatures can be associated with aging, leading to insights into
the genomic changes that accumulate over time [13].
In a specific example, Xu et al. developed a deep learning model to predict genomic
instability from histopathology slides. When applied to 1010 patients with breast cancer,
the model accurately classified chromosomal instability status with with 81.2% sensitivity
and 68.7% specificity in the test set [14].

2.2. Telomere Attrition

2.2.1. Mechanisms of Telomere Shortening and Its Impact on Cellular Aging
Telomeres are repetitive nucleotide sequences at the ends of chromosomes that pro-
tect them from degradation and prevent chromosomal fusions. With each cell division,
telomeres shorten due to the end-replication problem, where DNA polymerase cannot
fully replicate the ends of linear chromosomes. Critically short telomeres trigger a DNA
damage response, leading to cellular senescence or apoptosis. The enzyme telomerase
can elongate telomeres by adding telomeric repeats; however, its expression is limited in
most somatic cells. Telomere shortening has been implicated in the aging of high-turnover
tissues, such as hematopoietic and epithelial cells. Experimental studies have demon-
strated that telomere dysfunction activates the p53 pathway, leading to cell cycle arrest and
senescence. In contrast, overexpression of telomerase in mouse models extends lifespan
and delays the onset of age-related pathologies, highlighting the therapeutic potential of
targeting telomere attrition to combat aging [15].

2.2.2. Integrative Approaches to Study Telomere Dynamics Using Multi-Omics Data

ML models can integrate multi-omics datasets to analyze key regulators of telomere
dynamics and predict the effects of telomere shortening on cellular function. For example,
ML algorithms can predict the onset of cellular senescence based on the combination of
telomere length, gene expression profiles, and protein interaction networks. Furthermore,
ML can help stratify individuals based on their risk of telomere attrition-related conditions
by predicting telomere dynamics and linking them to specific age-related diseases [16].
J. Pers. Med. 2024, 14, 931 4 of 27

In a specific example, Zhang et al. developed an ML approach to predict telomere

length from whole-genome sequencing data. They used a random forest model trained on
features extracted from sequencing reads that mapped to telomeric regions. The model
was trained on and validated by data from the 1000 Genomes Project and showed high
accuracy in predicting telomere length. This method provides a cost-effective alternative to
experimental telomere length measurement and could be valuable for large-scale studies
on aging and age-related diseases [17].

2.3. Epigenetic Alterations

2.3.1. Age-Associated Changes in Epigenetic Markers
Epigenetic changes, including DNA methylation, histone modifications, and chro-
matin remodeling, play a crucial role in regulating gene expression. Age-associated epi-
genetic alterations can disrupt normal gene regulatory networks, leading to changes in
cellular function and contributing to aging and age-related diseases. For instance, global
DNA hypomethylation and site-specific hypermethylation of CpG islands in gene promot-
ers have been observed in aged tissues. These changes can lead to the silencing of tumor
suppressor genes and the activation of oncogenes, promoting cancer development. Histone
modifications, such as the loss of histone H3 lysine 27 trimethylation (H3K27me3), are also
associated with aging and contribute to the deregulation of gene expression. Research
using epigenetic clocks, which measure biological age based on DNA methylation patterns,
has shown that these clocks can predict lifespan and age-related disease risk, demonstrating
the significance of epigenetic changes in aging [18].

2.3.2. Machine Learning Techniques to Identify and Predict Epigenetic Alterations

Supervised ML models, such as linear regression and deep neural networks, can be
trained to predict biological age based on DNA methylation patterns, often referred to as
epigenetic clocks. These models are key in estimating an individual’s biological age versus
chronological age. Furthermore, ML can identify age-associated epigenetic markers by
clustering samples based on DNA methylation and histone modification data, providing a
clearer picture of how epigenetic alterations drive aging [19].
In a specific example, Higgins-Chen et al. introduced a new epigenetic clock using
deep neural networks, addressing key limitations of previous clocks. Their model, called
DunedinPACE, predicts the pace of aging from DNA methylation data. Unlike previous
clocks that estimate biological age, DunedinPACE focuses on the rate of aging, providing
a more dynamic measure. The model was trained on longitudinal data and validated
across multiple cohorts, showing improved reliability and sensitivity to biological aging
interventions [20].

2.4. Loss of Proteostasis

2.4.1. Disruption in Protein Homeostasis and Its Implications
Proteostasis, or protein homeostasis, involves the maintenance of the correct concen-
tration, conformation, and localization of proteins within cells. Disruption of proteostasis is
a hallmark of aging and is linked to various age-related diseases, including neurodegenera-
tive disorders like Alzheimer’s and Parkinson’s diseases. The accumulation of misfolded
and aggregated proteins, which escape the cellular quality control systems such as the
ubiquitin–proteasome system and autophagy, can lead to cellular toxicity and dysfunction.
Experimental models have shown that enhancing proteostasis pathways, such as upregulat-
ing chaperone proteins or activating autophagy, can ameliorate age-related proteotoxicity
and extend lifespan. For example, caloric restriction and pharmacological activation of
autophagy have been shown to reduce protein aggregation and improve cellular function
in aging models, highlighting potential therapeutic strategies for restoring proteostasis [21].
J. Pers. Med. 2024, 14, 931 5 of 27

2.4.2. Omics Data to Study Proteostasis and Machine Learning Models to Predict
Proteostasis-Related Disorders
ML models, especially deep learning approaches, can analyze large-scale proteomic
data to identify patterns of protein misfolding, aggregation, and degradation that are
characteristic of aging cells. Supervised learning can predict the likelihood of proteostasis-
related disorders, such as neurodegenerative diseases, by analyzing protein expression
profiles and identifying key proteins that lose stability with age [22,23].
In a specific example, Eshari et al. developed an logistic regression model to predict
protein aggregation propensity from sequence data. Their model outperformed existing
methods in predicting aggregation-prone regions in proteins. This approach could be
valuable for understanding protein misfolding in age-related neurodegenerative diseases
and for designing therapeutic interventions [24].

2.5. Disabled Macroautophagy

2.5.1. Importance of Autophagy in Aging and Age-Related Diseases
Macroautophagy, commonly referred to as autophagy, is a cellular process that de-
grades and recycles damaged organelles, proteins, and other macromolecules. Autophagy
is essential for maintaining cellular homeostasis and responding to stress. Impairment of
autophagy is associated with aging and age-related diseases, including neurodegenerative
disorders and cancers. Studies have shown that autophagy declines with age, leading to
the accumulation of damaged cellular components and increased cellular stress. Genetic
and pharmacological interventions that enhance autophagy have been shown to extend
lifespan and delay the onset of age-related pathologies in model organisms. For instance,
activation of the autophagy-related gene Atg5 in mice enhances autophagic activity and
promotes healthy aging, suggesting that targeting autophagy pathways could be a viable
strategy to combat aging [25].

2.5.2. Combining Multi-Omics Data to Understand and Enhance Autophagy Processes

ML can integrate transcriptomic, proteomic, and metabolomic data to predict the
levels of autophagic activity in cells. This can help identify individuals at risk for age-
related diseases due to impaired autophagy. By analyzing multi-omics data, ML models
can identify the key regulators of autophagy and suggest potential therapeutic targets to
restore autophagic processes in aging cells [26].
In a specific example, Dong et al. explored autophagy-related biomarkers in peripheral
blood for diagnosing rheumatoid arthritis using machine learning. Researchers identified
25 differentially expressed autophagy-related genes (DE-ARGs) and used algorithms like
random forest and LASSO to pinpoint three key biomarkers that showed strong diagnostic
value across multiple validation cohorts [27].

2.6. Deregulated Nutrient Sensing

2.6.1. Impact of Nutrient Sensing Pathways on Aging
Nutrient sensing pathways, such as the insulin/IGF-1 signaling (IIS) pathway, mecha-
nistic target of rapamycin (mTOR), AMP-activated protein kinase (AMPK), and sirtuins,
play critical roles in regulating metabolism, growth, and aging. Deregulation of these path-
ways can lead to metabolic imbalances and contribute to aging and age-related diseases.
For example, the mTOR pathway, which promotes anabolic processes and inhibits catabolic
processes like autophagy, is overactive in many age-related diseases. In contrast, activation
of AMPK and sirtuins, which promote catabolic processes and enhance stress resistance,
is associated with increased lifespan and improved metabolic health. Experimental evi-
dence from model organisms, such as caloric restriction and pharmacological inhibition of
mTOR, has demonstrated that the modulation of nutrient sensing pathways can extend
lifespan and delay the onset of age-related diseases, highlighting the therapeutic potential
of targeting these pathways [28].
J. Pers. Med. 2024, 14, 931 6 of 27

2.6.2. Integrative Models to Study the Regulation and Deregulation of These Pathways
Supervised ML algorithms can predict the activation state of nutrient sensing pathways
(like mTOR, AMPK, and sirtuins) based on gene expression and proteomic data, aiding
in the understanding of how these pathways contribute to aging. Unsupervised learning
techniques can identify novel regulatory networks within nutrient sensing pathways,
providing deeper insights into how these pathways are deregulated in aging [29].
In a specific example, Drewe et al. used machine learning models to identify AMPK
activators, which play a crucial role in regulating cellular metabolism and are beneficial in
treating diseases like diabetes and cancer. Various algorithms, including random forest and
deep neural networks, were tested and showed high accuracy in distinguishing activators
from controls. The models are suitable for screening potential AMPK activators, including
natural compounds, and can guide further in vitro testing to identify promising candidates
for drug development [30].

2.7. Mitochondrial Dysfunction

2.7.1. Role of Mitochondrial Function and Dysfunction in Aging
Mitochondria are the powerhouses of the cell, responsible for producing ATP through
oxidative phosphorylation. Mitochondrial dysfunction is a hallmark of aging and is charac-
terized by impaired energy production, increased production of reactive oxygen species
(ROS), and mitochondrial DNA (mtDNA) mutations. These changes contribute to cellular
damage and the decline in physiological functions observed in aging. Studies have shown
that mitochondrial biogenesis and dynamics, including fission and fusion processes, decline
with age, leading to the accumulation of dysfunctional mitochondria. Interventions that
enhance mitochondrial function, such as caloric restriction and mitochondrial-targeted
antioxidants, have been shown to extend lifespan and improve metabolic health in model
organisms, suggesting that targeting mitochondrial dysfunction could be a viable strategy
to combat aging [31].

2.7.2. Multi-Omics Approaches to Study Mitochondrial Health and Predictive Models

ML models can predict mitochondrial dysfunction by integrating various omics data
(genomic, transcriptomic, proteomic) and identifying patterns associated with impaired
mitochondrial function. ML can also reveal how mitochondrial dysfunction interacts with
other cellular processes, such as oxidative stress, by analyzing complex data interactions
and predicting the downstream effects [32].
In a specific example, Qin developed the mitochondrial programmed cell death index
as a strong prognostic tool for lower-grade glioma using ML on data from 1467 patients.
This index, based on 18 key genes, outperformed existing clinical models in predicting
patient outcomes. Patients with high index had worse survival, increased immune activ-
ity, and higher tumor mutation burden, suggesting its potential in guiding personalized
treatment and immunotherapy [33].

2.8. Cellular Senescence

2.8.1. Mechanisms and Consequences of Cellular Senescence
Cellular senescence is a state of permanent cell cycle arrest that occurs in response
to various stressors, such as DNA damage, oxidative stress, and telomere shortening.
Senescent cells contribute to aging and age-related diseases by secreting pro-inflammatory
factors and other molecules that disrupt tissue function, known as the senescence-associated
secretory phenotype (SASP). Studies have shown that the accumulation of senescent cells in
tissues leads to chronic inflammation, tissue dysfunction, and the promotion of age-related
diseases such as osteoarthritis and atherosclerosis. Experimental evidence from model
organisms has demonstrated that the selective elimination of senescent cells, known as
senolysis, can delay aging and extend lifespan, highlighting the therapeutic potential of
targeting cellular senescence [34].
J. Pers. Med. 2024, 14, 931 7 of 27

2.8.2. Using Omics Data and Machine Learning to Identify Senescent Cells and
Develop Interventions
Supervised ML models can classify cells based on their senescence status by analyzing
the expression of cell cycle inhibitors and SASP factors, allowing for the identification of
senescent cells within tissues. Furthermore, ML can also model the composition and dynam-
ics of the SASP, providing insights into how these factors contribute to tissue dysfunction
and aging [35].
In a specific example, Tuttle et al. conducted a systematic review and meta-analysis
of cellular senescence markers across various human tissues. While not directly applying
ML, their comprehensive analysis provides a foundation for future ML studies on cellular
senescence. They synthesized data from multiple studies to identify consistent markers of
senescence across different tissues and age groups. This work is crucial for developing ML
models to predict cellular senescence and its impact on aging [36].

2.9. Stem Cell Exhaustion

2.9.1. Decline in Stem Cell Function with Age
Stem cells are essential for tissue regeneration and repair. However, their function
declines with age, leading to impaired tissue maintenance and regeneration. Stem cell
exhaustion is characterized by a reduction in the number and function of stem cells,
which contributes to the decline in tissue homeostasis and repair capacity observed in
aging. Studies have shown that aged stem cells exhibit increased DNA damage, epigenetic
alterations, and a decline in their regenerative potential. Interventions that rejuvenate aged
stem cells, such as the activation of signaling pathways that promote stem cell self-renewal
or the transplantation of young stem cells, have been shown to improve tissue regeneration
and extend lifespan in model organisms, suggesting that targeting stem cell exhaustion
could be a viable strategy to combat aging [37].

2.9.2. Integrative Approaches to Study Stem Cell Biology and Predictive Models
ML models can predict the likelihood of stem cell exhaustion by analyzing gene
expression and protein interaction networks, helping to identify key factors that lead to the
decline in stem cell function with age. These models can also help in planning interventions
by predicting the effects of potential treatments on stem cell function, guiding strategies to
rejuvenate aged stem cells [38].
In a specific example, Barardo et al. developed an ML model to predict lifespan-
extending compounds, which could potentially address stem cell exhaustion. They used
a random forest classifier trained on various molecular and chemical features to predict
compounds that might extend lifespan. While not directly focused on stem cells, many
lifespan-extending compounds work by preserving stem cell function. This approach
demonstrates how ML can be used to identify potential interventions for age-related stem
cell exhaustion [39].

2.10. Altered Intercellular Communication

2.10.1. Changes in Cell Signaling and Communication in Aging
Cell signaling and communication are critical for maintaining tissue homeostasis.
Altered intercellular communication is a hallmark of aging and can lead to disrupted
tissue function and inflammation. For example, age-related changes in the immune system,
known as immunosenescence, can impair the body’s ability to respond to infections and
repair tissue damage. Studies have shown that the aged immune system exhibits a decline
in the production of cytokines and chemokines, leading to a reduced ability to mount an
effective immune response. Interventions that restore normal intercellular communication,
such as the administration of cytokines or the transplantation of young immune cells,
have been shown to improve immune function and extend lifespan in model organisms,
suggesting that targeting altered intercellular communication could be a viable strategy to
combat aging [40].
J. Pers. Med. 2024, 14, 931 8 of 27

2.10.2. Omics Data to Study Intercellular Communication and Machine Learning Models to
Predict Alterations
ML can classify cells based on their signaling activity by integrating transcriptomic
and proteomic data, helping to identify disruptions in cell communication that contribute
to aging. By revealing how altered intercellular communication interacts with processes
like inflammation, ML models can guide the development of therapies aimed at restoring
normal cell signaling [41].
In a specific example, Wang et al. developed iTALK, an R package that uses ML algo-
rithms to analyze single-cell RNA-seq data and predict cell–cell communication networks.
While not specifically focused on aging, this tool has significant potential for studying how
intercellular communication changes with age. iTALK can identify ligand-receptor pairs
and visualize communication networks, providing insights into how cellular interactions
are altered in aging tissues [42].

2.11. Chronic Inflammation

2.11.1. Role of Chronic Inflammation in Aging and Age-Related Diseases
Chronic inflammation, also known as “inflammaging”, is a characteristic of aging and
is linked to various age-related diseases, including CVD, diabetes mellitus, and neurode-
generative disorders. Persistent low-grade inflammation can contribute to tissue damage
and dysfunction, exacerbating the aging process. Studies have shown that aged tissues
exhibit increased levels of pro-inflammatory cytokines and a decline in anti-inflammatory
signaling pathways. Experimental evidence from model organisms has demonstrated that
the inhibition of pro-inflammatory pathways or the activation of anti-inflammatory path-
ways can extend lifespan and improve health span, highlighting the therapeutic potential
of targeting chronic inflammation [43].

2.11.2. Combining Multi-Omics Data to Study Inflammation and Predictive Models

ML models can classify cells based on their inflammatory activity by analyzing gene
expression profiles and proteomic data, identifying those with high levels of chronic
inflammation (inflammaging). These models can also reveal how chronic inflammation
interacts with other cellular processes, such as oxidative stress, offering insights into
potential anti-inflammatory interventions [44].
In a specific example, Bobrov et al. developed PhotoAgeClock, a deep learning model
that predicts biological age from facial images. While not directly measuring inflammation,
this non-invasive approach can capture signs of aging that are often related to chronic
inflammation. The model was trained on a large dataset of facial images and showed
high correlation with chronological age. This study demonstrates how ML can be used
to develop non-invasive biomarkers of aging, which could be valuable for studying the
effects of chronic inflammation on aging [45].

2.12. Dysbiosis
2.12.1. Age-Related Changes in the Microbiome
Dysbiosis refers to changes in the composition and function of the microbiome. The
microbiome, which consists of trillions of microorganisms residing in the gut and other
tissues, plays a critical role in regulating metabolism, immunity, and overall health. Alter-
ations in the microbiome can influence various aspects of health and contribute to aging
and age-related diseases. Studies have shown that aged individuals exhibit a decline in
microbial diversity and an increase in the abundance of pathogenic bacteria. Experimental
evidence from model organisms has demonstrated that the restoration of a healthy micro-
biome, through the administration of probiotics or fecal microbiota transplantation, can
improve metabolic health and extend lifespan, suggesting that targeting dysbiosis could be
a viable strategy to combat aging [46].
J. Pers. Med. 2024, 14, 931 9 of 27

2.12.2. Integrative Approaches to Study the Microbiome and Machine Learning Models to
Predict Dysbiosis
ML can classify microbiome samples based on their age-related profiles, using metage-
nomic and metabolomic data to identify signatures of dysbiosis that contribute to aging.
By integrating omics data, ML models can reveal how dysbiosis interacts with host pro-
cesses, such as metabolism and immune function, guiding the development of microbiome-
targeted interventions [46].
In a specific example, Wilmanski et al. (2021) used ML to analyze microbiome data and
predict biological age. They developed a random forest model trained on gut microbiome
composition data from a large cohort of individuals. The model identified specific microbial
features associated with healthy aging and longevity. The study demonstrated how ML
can be used to understand the complex relationship between the gut microbiome and
aging, providing insights into potential interventions to promote healthy aging through
microbiome modulation [47].
A summary of the aging hallmarks, their molecular mechanisms, and potential inter-
ventions is presented in Table 1.

Table 1. Summary of aging hallmarks, their molecular mechanisms, and potential interventions.

Aging Hallmark Molecular Mechanisms Examples of Interventions References

• PARP inhibitors (e.g., olaparib).

Increased frequency of mutations and
Genomic Instability • NAD+ precursors (e.g., nicotinamide riboside). [2,10–13,48,49]
structural alterations in DNA.
• Antioxidants (e.g., glutathione).

Telomere shortening due to end-replication • Telomerase activators (e.g., TA-65).

Telomere Attrition [2,15,16,50,51]
problem and limited telomerase expression. • Telomere protective compounds (e.g., resveratrol).

Changes in DNA methylation, histone • HDAC inhibitors (e.g., suberoylanilide hydroxamic acid).
Epigenetic Alterations [2,18,19,52,53]
modifications, and chromatin remodeling. • DNMT inhibitors (e.g., 5-azacytidine).

Accumulation of misfolded and • Chaperone activators (e.g., geldanamycin).

Loss of Proteostasis [2,21–23,54]
aggregated proteins. • Proteasome activators (e.g., spermidine).

Impairment of autophagy leads to the

• mTOR inhibitors (e.g., rapamycin).
Disabled Macroautophagy accumulation of damaged [2,25,26,55,56]
• AMPK activators (e.g., metformin).
cellular components.

Deregulated Nutrient Deregulation of insulin/IGF-1 signaling, • mTOR inhibitors (e.g., rapamycin).

[2,28,29,57]
Sensing mTOR, AMPK, and sirtuin pathways. • Sirtuin activators (e.g., resveratrol).

Impaired energy production, increased ROS, • Mitochondrial-targeted antioxidants (e.g., MitoQ).

Mitochondrial Dysfunction [2,31,32,58,59]
and mtDNA mutations. • PGC-1α activators (e.g., bezafibrate).

Permanent cell cycle arrest due to DNA • Senolytics (e.g., dasatinib, quercetin).
Cellular Senescence [2,34–36,60,61]
damage, oxidative stress, etc. • SASP inhibitors (e.g., ruxolitinib).

Reduction in the number and function of • Wnt pathway modulators (e.g., R-spondin1).
Stem Cell Exhaustion [2,37,38,62]
stem cells. • Young blood factors (e.g., GDF-11, TIMP-2).

Altered Intercellular • Cytokine modulators (e.g., tofacitinib).

Changes in cell signaling and communication. [2,40,41,63,64]
Communication • Immune rejuvenation therapies

Persistent low-grade inflammation is linked to • NF-κB inhibitors (e.g., aspirin).

Chronic Inflammation [2,43,44,65,66]
various age-related diseases. • Anti-inflammatory cytokines (e.g., IL-10).

Changes in the composition and function of • Probiotics (e.g., Lactobacillus spp.).

Dysbiosis [2,46,47,67]
the microbiome. • Fecal microbiota transplantation.

Abbreviations. AMPK (AMP-activated protein kinase); DNA (deoxyribonucleic acid); DNMT (DNA methyltrans-
ferase); GDF-11 (growth differentiation factor-11); HDAC (histone deacetylase); IGF-1 (insulin-like growth factor 1);
IL-10 (interleukin-10); mTOR (mechanistic target of rapamycin); NAD+ (nicotinamide adenine dinucleotide);
NF-κB (nuclear factor kappa-light-chain-enhancer of activated B cells); PARP (poly(ADP-ribose) polymerase);
PGC-1α (peroxisome proliferator-activated receptor gamma coactivator 1-alpha); ROS (reactive oxygen species);
SASP (senescence-associated secretory phenotype); TA-65 (telomerase activator-65); TIMP-2 (tissue inhibitor of
metalloproteinases-2); Wnt (wingless/integrated).
J. Pers. Med. 2024, 14, 931 10 of 27

3. Machine Learning Techniques in Medicine

3.1. Overview of Machine Learning Algorithms
ML encompasses a variety of algorithms that can be broadly categorized into super-
vised, unsupervised, and reinforcement learning.
In supervised learning, algorithms learn from labeled training data to make predic-
tions or classifications. Examples include linear regression, decision trees, random forests,
support vector machines (SVMs), and neural networks. In supervised learning, the algo-
rithm is trained on input–output pairs, allowing it to predict outcomes for new unseen
data [68].
In unsupervised learning, algorithms identify patterns or structures within unla-
beled data. Examples include clustering algorithms like k-means, hierarchical clustering,
and dimensionality reduction techniques like principal component analysis (PCA) and
t-distributed stochastic neighbor embedding (t-SNE). Unsupervised learning is useful
for discovering hidden patterns and structures in data without prior knowledge of the
outcomes [69].
In reinforcement learning, algorithms learn by interacting with an environment to
maximize cumulative rewards. This approach is often used in robotics, game playing, and
autonomous systems. Reinforcement learning involves learning a policy that maps states
to actions to achieve the highest reward [70].
In deep learning, algorithms use neural networks with multiple layers (hence “deep”)
to model complex patterns in large datasets. Examples include convolutional neural
networks (CNNs) for image data and recurrent neural networks (RNNs) for sequential
data. Deep learning models are particularly powerful for integrating multi-omics data,
identifying intricate patterns, and making accurate predictions in biomedical research [71].
In biomedical research, these algorithms can be applied to various tasks, including
disease diagnosis, treatment prediction, and biomarker discovery. For example, supervised
learning algorithms can predict patient outcomes based on clinical and omics data, while
unsupervised learning techniques can identify subgroups of patients with distinct molecu-
lar profiles [8]. Deep learning models can integrate various types of omics data to provide
comprehensive insights into complex biological processes and enhance predictive accuracy.

3.2. Challenges in Handling Biomedical Data

Biomedical data present several challenges, including data heterogeneity, integration,
and quality.
Data heterogeneity. Biomedical data come from various sources and formats, such as
genomic sequences, imaging data, and electronic health records. Integrating these diverse
datasets requires sophisticated data preprocessing and normalization techniques. For
example, genomic data may need to be aligned and variant-called, while imaging data may
require segmentation and feature extraction [72].
Data integration. Combining multi-omics data requires careful consideration of the dif-
ferent scales and types of data involved. Advanced computational models and algorithms
are needed to effectively integrate these datasets. For instance, multi-omics integration may
involve combining high-dimensional genomic data with lower-dimensional clinical data,
requiring techniques like matrix factorization or network-based approaches [73].
Data quality. Biomedical data can be noisy, incomplete, or biased. Ensuring data qual-
ity is crucial for the development of accurate and reliable ML models. Data preprocessing
steps, such as normalization, imputation, and outlier detection, are essential for improving
data quality. Additionally, techniques like cross-validation and bootstrapping can help
assess model robustness and reliability [71].
Solutions for these challenges include preprocessing techniques such as normalization
and imputation, feature selection methods to reduce dimensionality, and robust model
evaluation strategies to assess model performance. For example, normalization methods
can adjust for batch effects in omics data, while feature selection techniques can identify the
most informative variables for model building. Cross-validation and external validation
 techniques can identify the most informative variables for model building. Cro
validation and external validation on independent datasets are crucial for assessing mo
generalizability and preventing overfitting [69].
J. Pers. Med. 2024, 14, 931 11 of 27
3.3. Advances in Machine Learning for Multi-Omics Data Integration
Recent advances in ML have enabled the more sophisticated integration of mu
on independent datasets are crucial for assessing model generalizability and preventing
omicsoverfitting
data. Deep [69].learning techniques, such as convolutional neural networks (CNNs) a
recurrent neural networks (RNNs), have shown promise in analyzing complex biomedi
3.3. Advances in Machine Learning for Multi-Omics Data Integration
data. These models can automatically learn hierarchical representations from raw da
Recent advances in ML have enabled the more sophisticated integration of multi-
making them suitable for integrating multi-omics data [71].
omics data. Deep learning techniques, such as convolutional neural networks (CNNs) and
Graph-based
recurrent neural approaches,
networks (RNNs), such
haveasshown
graph neural
promise in networks (GNNs),
analyzing complex can capture
biomedical
relationships
data. These between
models can different omics layers
automatically and provide
learn hierarchical a more holistic
representations from rawview of biologi
data,
making
systems. themmethods
These suitable for integrating
can multi-omics data
model interactions [71]. genes, proteins, and metabolit
between
facilitatingGraph-based approaches,
the integration such as graph
of multi-omics neural
data networks (GNNs), can capture the
[74].
relationships between different omics layers and provide a more holistic view of biological
Transfer learning involves leveraging pre-trained models on large datasets
systems. These methods can model interactions between genes, proteins, and metabolites,
improve performance
facilitating on smaller
the integration task-specific
of multi-omics datasets. This approach can be particula
data [74].
useful forTransfer
integrating
learningmulti-omics data,pre-trained
involves leveraging where pretrained models
models on large canto provide
datasets improve valua
performance on smaller task-specific datasets.
insights into underlying biological processes [75]. This approach can be particularly useful for
integrating multi-omics data, where pretrained models can provide valuable insights into
Explainable AI techniques aim to provide interpretable and transparent ML mode
underlying biological processes [75].
These methods
ExplainablecanAIhelp researchers
techniques understand
aim to provide the and
interpretable decision-making process of M
transparent ML models.
models,
These ensuring
methods canthat the results
help researchers are the
understand biologically meaningful
decision-making process of MLand trustwort
models,
ensuring that the results are biologically meaningful and trustworthy. Explainable
Explainable artificial intelligence (AI) can enhance the interpretability of multi-omics d artificial
intelligence (AI) can enhance the interpretability of multi-omics data integration and guide
integration and guide the development of personalized interventions [76].
the development of personalized interventions [76].
The The
workflow
workflow ofofmulti-omics data
multi-omics data integration
integration withwith
ML inML in research
aging aging research is illustra
is illustrated
in Figure 1.
in Figure 1.

Figure
Figure 1. Workflow
1. Workflow ofofmulti-omics
multi-omics data
dataintegration withwith
integration machine learning
machine in aging research.
learning in aging research.
A summary of ML techniques and their applications in multi-omics aging research is
A summary
presented of ML
in Table 2. techniques and their applications in multi-omics aging research
presented in Table 2.
J. Pers. Med. 2024, 14, 931 12 of 27

Table 2. Summary of machine learning techniques and their applications in multi-omics aging research.

ML Technique Description Applications

• Predicting genomic instability.

• Classifying mutations as benign or pathogenic.
Algorithms that learn from labeled data to • Predicting biological age.
Supervised Learning
make predictions. • Identifying aging biomarkers.
• Disease risk prediction.

• Discovering new age-associated biomarkers.

• Identifying distinct epigenetic signatures.
Algorithms that identify patterns in • Clustering age-related gene expression profiles and
Unsupervised Learning
unlabeled data.
proteomic changes.
• Grouping proteins based on stability.

Algorithms that learn by interacting with an • Optimizing intervention strategies.

Reinforcement Learning • Personalized treatment plans.
environment to maximize cumulative rewards.

• Integrating multi-omics data, predicting age-related

Neural networks with multiple layers can model disease progression.
Deep Learning • Learning hierarchical representations from raw data.
complex relationships in data.
• Improving predictive accuracy for age-related diseases.

Models that capture relationships and • Modeling interactions between genes, proteins,
Graph Neural Networks interactions between entities in a and metabolites.
graph structure. • Integrating multi-omics data for comprehensive analysis.

Convolutional Neural Deep learning models are particularly effective • Integrating and analyzing imaging data.
Networks for analyzing spatial and visual data. • Detecting patterns in omics data.

• Enhancing model accuracy by leveraging

Leveraging pretrained models to improve pre-trained models.
Transfer Learning
performance on new smaller datasets. • Transferring knowledge across different datasets.

• Providing interpretable and transparent machine

Explainable Artificial Techniques to make artificial intelligence models’ learning models.
Intelligence decisions interpretable and transparent. • Ensuring results are biologically meaningful, enhancing
trust and understanding of machine learning predictions.

3.4. Practical Guidance on Selecting Machine Learning Methods for Multi-Omics Research
in Geroscience
Data from multi-omics technologies are often characterized by diversity and com-
plexity, hindering analysis and drawing conclusions. When selecting ML methods for
multi-omics integration in geroscience, it is crucial to consider the specific characteristics of
the data and the research objectives. As a practical guide, we can apply the following ML
methods for each of the following basic indications [20,77–80]:
• Genomic, epigenomic, and proteomic data analysis.
- Random forests: ideal for telomere length analysis.
- Support vector machines (SVMs): Suitable for distinguishing between benign and
pathogenic mutations in genomic instability studies. Also ideal for identification
of proteins with high aggregation potential.
- Gradient boosting machines (GBMs): effective in identifying key epigenetic modi-
fications that contribute to aging.
- CNNs: best for analyzing proteomic data, particularly when dealing with struc-
tural data like protein imaging or spatial transcriptomics, where the spatial rela-
tionships between features are important.
- Recurrent neural networks (RNNs) or long short-term memory (LSTM): Ideal for
analyzing genomic sequences where the order of nucleotides (sequential data) is
crucial. These models are particularly effective in understanding mutations that
affect protein structure and function.
J. Pers. Med. 2024, 14, 931 13 of 27

- Fully connected neural networks (FCNNs): suitable for integrating multi-omics

data (e.g., genomic, epigenomic, and proteomic) to classify complex age-related
changes where high-dimensional data need to be processed.
• Longitudinal or dynamic data analysis.
- RNNs or long short-term memory (LSTM): Particularly suited for analyzing
changes in biomarkers over time. These networks are designed to handle sequen-
tial data, making them ideal for capturing temporal patterns in longitudinal data.
• Predictive models.
- Ensemble methods: useful for robust predictive modeling by combining the
strengths of multiple algorithms.
- Multiple regression combined with ML techniques: offers a simpler interpretable
approach to prediction when the focus is on a few key variables.
• Novel biomarker discovery.
- CNNs: ideal when the biomarker discovery involves image data or spatially
structured data, such as histopathological images or spatial transcriptomics data.
- Autoencoders: A type of neural network ideal for unsupervised learning. Ideal
when we have large complex multi-omics data and we want to find hidden
patterns, making it easier to spot new biomarkers.
• Drug efficacy predictions and personalized medicine interventions.
- GBMs.
• Modeling complex interactions among various age-related pathways (e.g., autophagy).
- deep learning models, including GNNs for graph-based approaches.
• Explainable AI techniques: should be incorporated to ensure that the outcomes of ML
models are interpretable and actionable, particularly in a clinical setting where the
biological significance of the results must be clearly understood.

4. Case Studies and Applications

4.1. Predictive Modeling of Age-Related Diseases
ML models have been successfully applied to predict age-related diseases by inte-
grating multi-omics data. These models can identify individuals at high risk and guide
personalized interventions [48]. Some notable examples include Alzheimer’s disease (AD)
and CVD.
AD is a neurodegenerative disorder characterized by cognitive decline and memory
loss. Integrating genomic, transcriptomic, and proteomic data has led to the identifica-
tion of biomarkers and predictive models for AD. For instance, ML models have been
used to identify genetic variants and gene expression profiles associated with AD risk.
Proteomic analyses have revealed changes in protein expression and post-translational
modifications that correlate with disease progression. Integrating these datasets with
clinical data has improved the accuracy of predictive models and facilitated early diag-
nosis and intervention [49]. AI and ML have also optimized diagnostic procedures and
outperformed traditional neuropsychological tests in identifying cognitive impairment.
For example, AI-driven mobile screening tests and game-based intelligence tests have
demonstrated superior performance compared with conventional methods. Additionally,
AI techniques have enhanced virtual reality assessments, offering high ecological validity
and new avenues for cognitive rehabilitation [50]. A recent study proposed an ML model
trained on a set of neuropsychological, neurophysiological, and clinical data to predict
cognitive decline in MCI and AD patients. The study collected data from 4848 patients,
including those diagnosed with AD and MCI. The results showed a diagnostic accuracy
of 86%, with a sensitivity of 72% and a specificity of 91% for clinical data prediction with
MMSE scores [51].
CVD is a leading cause of mortality worldwide. Integrating multi-omics data has
provided valuable insights into the molecular mechanisms underlying CVD. For exam-
J. Pers. Med. 2024, 14, 931 14 of 27

ple, genomic data can identify genetic predispositions to CVD, while transcriptomic and
proteomic data can reveal changes in gene and protein expression associated with disease
states. Metabolomic profiling can provide information on metabolic shifts that contribute
to CVD. ML models have been used to integrate these datasets, leading to the development
of predictive models for CVD risk and progression. A recent study has utilized ML to
investigate the presence of multiple factors in the risk of aortic stenosis. The findings
indicated that significant features present in aortic stenosis patients included older age,
arterial hypertension, aortic regurgitation, ascending aortic dilatation, and bicuspid aortic
valve. These insights suggest that hypertension and other factors play a crucial role in the
hemodynamic and anatomical progression of AS, highlighting the implications of ML for a
comprehensive risk assessment and early intervention strategy [52].

4.2. Personalized Medicine Approaches for Aging Populations

4.2.1. Genomic Instability Interventions
Targeting genomic instability involves enhancing DNA repair mechanisms and protect-
ing against DNA damage. Several interventions have shown promise in preclinical models.
Poly(ADP-ribose) polymerase (PARP) inhibitors. These compounds enhance the
efficacy of DNA repair by inhibiting PARP enzymes, which are involved in the repair of
single-strand breaks. Studies in mice have demonstrated that PARP inhibitors can reduce
DNA damage and extend lifespan. ML models can identify individuals with high levels of
DNA damage or mutations that may benefit from PARP inhibitors by analyzing genomic
and epigenomic data to pinpoint those most likely to respond positively [53].
NAD+ precursors. Nicotinamide riboside (NR) and nicotinamide mononucleotide
(NMN) are precursors of NAD+, a coenzyme involved in DNA repair and cellular metabolism.
Supplementation with NAD+ precursors has been shown to improve DNA repair capacity,
reduce genomic instability, and extend lifespan in mice. ML can analyze metabolomic
profiles to identify candidates with low NAD+ levels or impaired NAD+ metabolism who
may benefit from such supplementation [54].

4.2.2. Telomere Attrition Interventions

Telomere attrition can be addressed by enhancing telomerase activity or protecting
telomeres from damage.
Telomerase activators. Compounds that activate telomerase, such as TA-65, have
been shown to elongate telomeres and improve cellular function in aged mice. These
interventions can delay the onset of age-related diseases and extend lifespan. ML models
can predict telomere length and assess telomerase activity from genomic and proteomic
data, identifying individuals who would benefit most from telomerase activators [81].
Telomere protective compounds. Certain small molecules and natural compounds,
such as resveratrol and its derivatives, have been shown to protect telomeres from oxidative
damage and improve telomere maintenance. These compounds can enhance telomere
stability and promote healthy aging. By analyzing oxidative stress markers and telomere
length data, ML can identify individuals with significant telomere attrition and oxidative
damage who may benefit from telomere-protective compounds [55].

4.2.3. Epigenetic Alterations Interventions

Targeting epigenetic alterations involves modulating DNA methylation and histone
modifications to restore youthful gene expression patterns.
DNA methyltransferase (DNMT) inhibitors. Compounds that inhibit DNMTs, such as
5-azacytidine, can reduce aberrant DNA methylation and restore normal gene expression
patterns. Preclinical studies have shown that DNMT inhibitors can improve metabolic
health and extend lifespan in mice. ML algorithms can analyze epigenomic data to detect
aberrant methylation patterns and predict which individuals might benefit from DNMT
inhibitors [56].
J. Pers. Med. 2024, 14, 931 15 of 27

Histone deacetylase (HDAC) inhibitors. HDAC inhibitors, such as suberoylanilide

hydroxamic acid (SAHA), can enhance histone acetylation and promote a more open
chromatin structure. These compounds have been shown to improve cognitive function and
extend lifespan in animal models. ML models can predict beneficial epigenetic changes and
identify candidates with histone modification profiles indicative of age-related epigenetic
alterations [57].

4.2.4. Loss of Proteostasis Interventions

Restoring proteostasis involves enhancing protein quality control systems and pro-
moting the degradation of misfolded proteins.
Chaperone activators. Small molecules that activate heat shock proteins (HSPs), such
as geldanamycin, can enhance protein folding and reduce the accumulation of misfolded
proteins. Preclinical studies have demonstrated that chaperone activators can improve
proteostasis and extend lifespan in mice. ML can analyze proteomic data to identify
individuals with high levels of misfolded proteins who might benefit from chaperone
activators [21].
Proteasome activators. In addition to chaperone activators, proteasome activators rep-
resent another promising avenue for restoring proteostasis in aging cells. The proteasome
is a complex protein assembly responsible for degrading damaged or misfolded proteins
tagged with ubiquitin, a process vital for maintaining cellular function and preventing
disease. Enhancing proteasome activity has been shown to alleviate proteotoxic stress and
improve cellular resilience against aging-related dysfunctions. For example, studies have
demonstrated that compounds like bortezomib, originally used in cancer therapy to induce
proteasome inhibition, can be tuned to subtly activate proteasome activity, thus promoting
the clearance of toxic proteins in neurodegenerative disease models [58]. Implications
of ML can significantly augment the development and application of proteasome activa-
tors in aging research. By analyzing vast datasets of proteomic profiles, ML algorithms
can identify patterns indicating proteasome inefficiency and predict which individuals or
models might benefit most from targeted proteasome activation. Furthermore, ML can
assist in screening for novel proteasome activators by predicting their efficacy and safety
based on structural and functional data. This predictive capability is crucial for optimizing
therapeutic strategies and tailoring interventions to specific aging populations or disease
phenotypes, thus embodying the principles of personalized medicine.

4.2.5. Disabled Macroautophagy Interventions

Enhancing macroautophagy involves stimulating the autophagic machinery to clear
damaged cellular components.
mTOR inhibitors. Compounds that enhance autophagy, such as mTOR inhibitors, can
promote the degradation of damaged proteins and organelles. Preclinical studies have
shown that mTOR inhibition can enhance autophagy, improve metabolic health, reduce
protein aggregation, and extend lifespan in mice. ML models can assess autophagic activity
from multi-omics data and identify those with impaired autophagy who could benefit from
these enhancers [59]. Furthermore, ML can predict mTOR pathway activity by analyzing
transcriptomic and proteomic data and identifying individuals with overactive mTOR
signaling who might benefit from mTOR inhibitors [60].
AMPK activators. Metformin and other AMPK activators stimulate autophagy by
activating the energy-sensing AMPK pathway. These compounds have been shown to
improve insulin sensitivity, reduce inflammation, and extend lifespan in animal models.
ML models can integrate metabolomic and transcriptomic data to identify individuals with
low AMPK activity who might benefit from AMPK activators [61].

4.2.6. Deregulated Nutrient Sensing Interventions

Modulating nutrient sensing pathways involves targeting key regulators such as
mTOR, AMPK, and sirtuins.
J. Pers. Med. 2024, 14, 931 16 of 27

mTOR inhibitors. As mentioned earlier, rapamycin and its analogs inhibit mTOR
signaling, which can mimic the effects of caloric restriction and promote longevity. Pre-
clinical studies have demonstrated that mTOR inhibitors can extend lifespan and improve
metabolic health in various animal models. ML can identify individuals with overactive
mTOR signaling through integrated omics data analysis, suggesting those who may benefit
from such interventions [30].
Sirtuin activators. Resveratrol and other sirtuin activators enhance the activity of
sirtuins, a family of NAD+-dependent deacetylases involved in metabolic regulation and
stress resistance. These compounds have been shown to improve mitochondrial function,
enhance stress resistance, and extend lifespan in animal models. ML can analyze sirtuin
activity from proteomic and metabolomic data to identify candidates who would benefit
from sirtuin activators [62].

4.2.7. Mitochondrial Dysfunction Interventions

Improving mitochondrial function involves enhancing mitochondrial biogenesis and
protecting against oxidative damage.
Mitochondrial-targeted antioxidants. Compounds such as MitoQ and SkQ1 selec-
tively target mitochondria to reduce oxidative stress. Preclinical studies have shown that
mitochondrial-targeted antioxidants can improve mitochondrial function, reduce ROS pro-
duction, and extend lifespan in animal models. ML can predict mitochondrial dysfunction
by analyzing mtDNA mutations and oxidative stress markers, identifying individuals who
might benefit from mitochondrial-targeted antioxidants [63].
PPARγ coactivator 1α (PGC-1α) activators. PGC-1α is a key regulator of mitochondrial
biogenesis. Compounds that activate PGC-1α, such as bezafibrate, have been shown to
enhance mitochondrial function and extend lifespan in mice. ML can assess PGC-1α
activity from transcriptomic and proteomic data to identify candidates with impaired
mitochondrial biogenesis who might benefit from these activators [64].

4.2.8. Cellular Senescence Interventions

Targeting cellular senescence involves selectively eliminating senescent cells or modu-
lating the senescence-associated secretory phenotype (SASP).
Senolytics. Compounds such as dasatinib and quercetin selectively induce apoptosis
in senescent cells. Preclinical studies have shown that senolytics can reduce the burden
of senescent cells, improve tissue function, and extend lifespan in mice. ML can identify
biomarkers of cellular senescence from multi-omics data, pinpointing individuals who
would benefit from senolytic therapies [65].
SASP inhibitors. JAK inhibitors, such as ruxolitinib, can modulate the pro-inflammatory
secretome of senescent cells. These compounds have been shown to reduce chronic inflam-
mation, improve tissue function, and extend lifespan in animal models. ML can predict the
SASP profile from transcriptomic and proteomic data, identifying those with high SASP
activity who might benefit from SASP inhibitors [66].

4.2.9. Stem Cell Exhaustion Interventions

Rejuvenating stem cells involves enhancing stem cell function and promoting their
self-renewal capacity.
Wnt pathway modulators. Activation of the Wnt signaling pathway has been shown
to enhance stem cell self-renewal and improve tissue regeneration. Compounds such as
R-spondin1 have been shown to rejuvenate aged stem cells and extend lifespan in mice.
ML can analyze stem cell markers and Wnt signaling activity from multi-omics data to
identify candidates who might benefit from Wnt pathway modulators [37].
Young blood factors. Transfusion of blood from young to old animals has been shown
to rejuvenate aged tissues and improve stem cell function. Factors such as GDF11 and
TIMP2 have been identified as potential mediators of these rejuvenating effects. ML can
J. Pers. Med. 2024, 14, 931 17 of 27

predict the benefits of young blood factors by analyzing circulating biomarkers and stem cell
function data, and identifying individuals who might benefit from such interventions [67].

4.2.10. Altered Intercellular Communication Interventions

Restoring normal intercellular communication involves modulating signaling path-
ways and improving immune function.
Cytokine modulators. IL-6 inhibitors, such as tocilizumab, can reduce pro-inflammatory
signaling and improve immune function. Preclinical studies have shown that cytokine
modulators can reduce chronic inflammation and extend lifespan in animal models. ML
can analyze cytokine profiles and immune function data to identify individuals with dys-
regulated intercellular communication who might benefit from cytokine modulators [82].
Immune rejuvenation. Interventions that rejuvenate the immune system, such as
thymic regeneration or the transplantation of young immune cells, have been shown to
improve immune function and extend lifespan in animal models. ML can predict immune
system aging from multi-omics data, identifying candidates for immune rejuvenation
therapies [83].

4.2.11. Chronic Inflammation Interventions

Reducing chronic inflammation involves targeting pro-inflammatory pathways and
enhancing anti-inflammatory signaling.
NF-κB inhibitors. Compounds that inhibit the NF-κB pathway, such as aspirin and sal-
icylates, can reduce chronic inflammation and improve metabolic health. Preclinical studies
have demonstrated that NF-κB inhibitors can extend lifespan and improve health span in
animal models. ML can identify individuals with high NF-κB activity from transcriptomic
and proteomic data, suggesting those who might benefit from NF-κB inhibitors [84].
Anti-inflammatory cytokines. IL-10 and other anti-inflammatory cytokines can modu-
late the immune response and reduce chronic inflammation. These cytokines have been
shown to improve tissue function and extend lifespan in preclinical studies. ML can
predict the benefits of anti-inflammatory cytokines by analyzing inflammatory markers
and immune function data and identifying individuals who might benefit from these
therapies [85].

4.2.12. Dysbiosis Interventions

Restoring a healthy microbiome involves modulating the composition and function of
the gut microbiota.
Probiotics. The administration of beneficial bacteria, such as Lactobacillus and Bi-
fidobacterium, can improve gut health and metabolic function. Preclinical studies have
shown that probiotics can extend lifespan and improve health span in animal models. ML
can analyze microbiome composition and metabolic profiles to identify individuals with
dysbiosis who might benefit from probiotic interventions [46].
Fecal microbiota transplantation (FMT). Transplanting gut microbiota from young
to old animals has been shown to rejuvenate the gut microbiome and improve metabolic
health. Preclinical studies have demonstrated that FMT can extend lifespan and improve
health span in animal models. ML can predict the benefits of FMT by analyzing micro-
biome diversity and function data and identifying individuals who might benefit from this
intervention [86].

4.3. Identification of Novel Biomarkers

The discovery of novel biomarkers is crucial for early diagnosis and treatment of
age-related diseases. ML techniques can be used to analyze multi-omics data and identify
potential biomarkers. These biomarkers can then be validated and their efficacy predicted
using ML models, providing valuable tools for clinical applications.
J. Pers. Med. 2024, 14, 931 18 of 27

4.3.1. Biomarker Discovery in Neurodegenerative Diseases

Neurodegenerative diseases, such as Parkinson’s disease and amyotrophic lateral
sclerosis, are characterized by progressive neuronal loss and functional decline. Identifying
biomarkers for early diagnosis and monitoring disease progression is challenging due to the
complex and multifactorial nature of these diseases. Integrating multi-omics data, includ-
ing genomic, transcriptomic, proteomic, and metabolomic profiles, can reveal molecular
changes associated with neurodegeneration. ML models can identify key biomarkers that
distinguish between different disease stages and predict disease progression, facilitating
early intervention and personalized treatment [48].

4.3.2. Biomarker Discovery in Inflammatory Diseases

Inflammatory diseases, such as rheumatoid arthritis and inflammatory bowel dis-
ease, involve dysregulated immune responses and chronic inflammation. Identifying
biomarkers for early diagnosis and monitoring disease activity is essential for effective
treatment. Integrating multi-omics data, including genomic, transcriptomic, proteomic,
and metabolomic profiles, can reveal molecular changes associated with inflammation. ML
models can identify key biomarkers that predict disease activity and treatment response,
guiding personalized interventions and improving patient outcomes [87].

4.3.3. Biomarker Discovery in Reproductive Aging

Recent studies have leveraged multi-omics approaches to identify potential biomarkers
for ovarian aging. One notable discovery is the multifunctional protein secreted phospho-
protein 1 (SPP1), also known as osteopontin. This protein has been implicated in various
biological processes, including inflammation, immune responses, and tissue remodeling.
However, its role in ovarian aging has not been fully explored until now. Spatial transcrip-
tomic analyses of mouse ovaries have revealed a significant decline in SPP1 expression
in aging tissues compared with younger ones. Furthermore, single-cell RNA sequencing
highlighted associations between SPP1 and key genes such as ITGAV, ITGB1, CD44, MMP3,
and FN1, indicating its regulatory role in the ovarian microenvironment. Additionally, data
from the Human Protein Atlas showed a marked decrease in SPP1 levels in the ovaries of
older women (aged 40–49) compared with younger women (aged 20–29). These findings
suggest that SPP1 could serve as a valuable biomarker for ovarian aging, aiding in the early
diagnosis and personalized treatment of age-related reproductive conditions. The decline in
SPP1 expression with age underscores its potential utility in developing targeted therapies
aimed at enhancing reproductive health and managing ovarian aging more effectively [88].

4.3.4. Biomarker Discovery in Cancer Research

Recent advancements in multi-omics and ML have also enhanced biomarker identifi-
cation for malignancies, including non-small-cell lung cancer. A notable study integrated
RNA-Seq, miRNA-Seq, copy number variation, and DNA methylation data with whole-
slide imaging data to identify potential biomarkers.
ML models were trained on each data type independently and their outputs were
combined using a late fusion approach, significantly improving diagnostic accuracy. This
method provided robust biomarkers that effectively distinguished between NSCLC sub-
types, demonstrating high-performance metrics like F1 score and AUC. The study also
highlighted the potential for integrating these biomarkers into clinical decision support
systems, enhancing personalized diagnostic and treatment strategies for lung cancer [89].

4.4. Applications of Machine Learning Algorithms in Medical Diagnosis

Recent developments in ML techniques have shown significant promise in enhancing
the analysis and integration of multi-omics data, particularly in medical diagnostics. For
example, a novel approach discussed in the recent literature involves a hybrid model that
combines CNNs with a pruned ensemble of extreme learning machines (ELMs). This
model, designed for breast cancer detection and diagnostics, leverages the strengths of
J. Pers. Med. 2024, 14, 931 19 of 27

both deep learning and ML to improve image analysis accuracy to 86%, surpassing tradi-
tional methods. By integrating CNN for robust feature extraction and ELM for efficient
classification, this model exemplifies the potential of hybrid ML systems in not only en-
hancing diagnostic accuracy but also in facilitating early disease detection, which is critical
in treatment planning and outcome improvement [90].
Additionally, AI is increasingly playing a transformative role in cardiology, enabling
enhanced diagnostic and predictive capabilities. ML algorithms, particularly neural net-
works, have been applied successfully in interpreting complex cardiovascular imaging
data, leading to improved diagnostic accuracy in conditions such as coronary artery dis-
ease, hypertrophic cardiomyopathy, and atrial fibrillation. These AI-driven tools assist
clinicians by providing more precise and rapid interpretations of echocardiograms and
electrocardiograms, potentially leading to better patient outcomes by enabling timely and
accurate treatment decisions [91]. Specifically, a recent study has explored the application
of ML in predicting atrial fibrillation in patients with embolic strokes of undetermined
sources (ESUS), showcasing another promising dimension of AI in cardiovascular health
diagnostics. A study utilized ML models, including SVMs, multilayer perceptron (MLP),
XGBoost, and random forest, to analyze clinical and echocardiographic data of 157 ESUS
patients. The SVM model exhibited the highest efficacy, with an area under the curve (AUC)
of approximately 0.736, demonstrating a robust capability to predict AF occurrences [92].
Another study focused on predicting myocardial ischemia by integrating exploratory data
analysis and ML models, achieving accuracies above 80% [93].
Furthermore, recent advancements in ML have significantly enhanced prostate cancer
diagnostics, particularly through the use of radiomics and AI in molecular imaging. By ap-
plying data-characterization algorithms to positron emission tomography scans, radiomics
helps to accurately distinguish between pathological and physiological tracer uptakes.
ML models, including CNNs, can analyze these radiomics features to predict Gleason
scores, influencing treatment decisions and enhancing personalized patient management
in oncology [94].
Such innovative approaches highlight the transformative impact of advanced ML
techniques in medical diagnosis, offering new pathways for the effective analysis of complex
biological datasets.

5. Challenges and Limitations

5.1. Data Availability, Integration, and Computational Complexity
A limitation of integrating ML with multi-omics technologies is the assumption of
readily available large well-annotated omics datasets. In the real world, such comprehen-
sive datasets may be scarce, particularly in underfunded research environments or regions
with limited technological infrastructure. Even in the presence of such data, integrating
diverse datasets from different omics layers presents significant computational challenges.
Advanced computational models and algorithms are required to handle the complexity
of multi-omics data integration. Additionally, the development of user-friendly tools and
platforms is essential to facilitate the application of these models in biomedical research [95].

5.2. Data Heterogeneity

Biomedical data come from various sources and formats, such as genomic sequences,
imaging data, and electronic health records. Integrating these diverse datasets requires
sophisticated data preprocessing and normalization techniques. For example, genomic data
may need to be aligned and variant-called, while imaging data may require segmentation
and feature extraction [95].

5.3. Data Quality

Biomedical data can be noisy, incomplete, or biased. Ensuring data quality is crucial
for the development of accurate and reliable ML models. Data preprocessing steps, such as
normalization, imputation, and outlier detection, are essential for improving data quality.
J. Pers. Med. 2024, 14, 931 20 of 27

Additionally, techniques like cross-validation and bootstrapping can help assess model
robustness and reliability [96].

5.4. Ethical Considerations and Data Privacy

The use of multi-omics data in biomedical research raises several ethical considerations
and data privacy concerns.

5.4.1. Informed Consent

Researchers must obtain informed consent from participants for the collection, storage,
and analysis of their biological samples and data. Participants should be informed about
the purpose of the study, the types of data being collected, and the potential risks and
benefits [97].

5.4.2. Data Privacy and Security

Protecting the privacy and confidentiality of participants’ data is paramount. Strategies
to ensure data privacy include data anonymization, secure data storage, and controlled
access to data. Implementing robust data security measures, such as encryption and secure
data transfer protocols, is also essential [98].

5.4.3. Ethical Use of Data

Researchers must ensure that the data are used ethically and responsibly. This includes
avoiding misuse or misinterpretation of data, maintaining transparency in data analysis
and reporting, and addressing potential biases in data collection and analysis [99].

5.4.4. Equity and Inclusion

Ensuring that diverse populations are represented in biomedical research is critical for
generalizability and equity. Researchers should strive to include participants from different
demographic backgrounds, including age, gender, ethnicity, and socioeconomic status, to
capture a comprehensive understanding of biological processes and their implications for
different populations [100].

5.5. Limitations of Current Studies

Current studies on aging often have limitations, such as small sample sizes, lack of
longitudinal data, and potential biases. Small sample sizes can limit the statistical power
and generalizability of study findings. Collaborative efforts to pool data from multiple
studies can help increase sample sizes and improve the robustness of results. Longitudinal
data are essential for studying the dynamic processes of aging and disease progression.
However, collecting longitudinal data can be challenging due to the time and resources
required. Developing standardized protocols for longitudinal data collection and analysis
can help address this challenge. Potential biases in data collection and analysis can impact
the validity of study findings. Addressing these biases requires careful consideration
of study design, data collection methods, and statistical analysis techniques. Ensuring
diversity and inclusion in study populations can also help mitigate biases.
Addressing these limitations requires collaborative efforts in data sharing, the inclu-
sion of diverse populations in research, and the development of standardized protocols for
data collection and analysis [95].

6. Future Directions
6.1. Advances in Omics Technologies
Emerging omics technologies, such as single-cell omics and spatial transcriptomics,
have the potential to provide unprecedented insights into the aging process. These tech-
nologies can capture molecular changes at higher resolution and across different tissues,
offering new opportunities for aging research.
J. Pers. Med. 2024, 14, 931 21 of 27

6.1.1. Single-Cell Omics

Single-cell genomics and transcriptomics, including single-cell RNA sequencing
(scRNA-seq) and single-cell ATAC sequencing (scATAC-seq), enable the analysis of gene
expression and chromatin accessibility at the single-cell level. Single-cell proteomics can
analyze protein levels and post-translational modifications in individual aging cells. These
technologies can reveal cellular heterogeneity and identify rare cell populations involved in
aging and age-related diseases. Integrating single-cell omics data with other omics layers
can provide a comprehensive view of the molecular mechanisms underlying aging [101].

6.1.2. Spatial Omics

This approach can map gene and protein expression in tissues, maintaining the spatial
context that is crucial for understanding the complex cellular interactions during aging.
It could lead to discoveries about how tissue microenvironments change with age and
contribute to senescence or regeneration [102].

6.2. Development of Sophisticated Machine Learning Models

Innovations in ML, such as deep learning and explainable AI, can further enhance the
analysis of multi-omics data. Collaborative efforts in data sharing and open science will be
crucial for advancing the field and translating research findings into clinical applications.

6.2.1. Deep Learning

Deep learning techniques, such as convolutional neural networks and recurrent neural
networks, have shown promise in analyzing complex biomedical data. These models can
automatically learn hierarchical representations from raw data, making them suitable for
integrating multi-omics data. Neural networks can model complex patterns and interac-
tions within large-scale omics data, potentially identifying new biomarkers of aging or
targets for drug development [71].

6.2.2. Explainable AI
Explainable AI techniques aim to provide interpretable and transparent ML models.
These methods can help researchers understand the decision-making process of ML models,
ensuring that the results are biologically meaningful and trustworthy. Explainable AI can
enhance the interpretability of multi-omics data integration and guide the development
of personalized interventions [76]. This is particularly important in precision medicine,
where the ability to interpret and trust AI models can significantly impact patient outcomes.
Explainable AI not only supports the integration of AI with digital health data but also
helps in improving the accuracy and reliability of predictive models, ensuring that the
results are actionable and meaningful for clinical practice [103].

6.3. Implementation in Personalized Medicine

The hallmark-targeted interventions discussed above hold significant promise for
future personalized medicine approaches. These approaches targeting the hallmarks of
aging hold the potential to revolutionize aging research and extend health span and lifespan.
Implementing these personalized interventions requires advanced computational tools
and ML models to analyze and integrate multi-omics data, predict individual responses
to treatments, and optimize therapeutic strategies. ML can provide insights into the
molecular profiles of patients, identify potential candidates for specific interventions,
and continuously refine treatment plans based on ongoing data collection and analysis.
Specifically, by integrating ML techniques with multi-omics data, healthcare providers can
perform the following:
• Identify biomarkers (e.g., proteins, RNAs, and metabolites) and genetic risk factors
that predict the onset of age-related diseases (e.g., Alzheimer’s or CVD) well before
clinical symptoms manifest.
 insights into the molecular profiles of patients, identify potential candidates for specific
interventions, and continuously refine treatment plans based on ongoing data collection
and analysis. Specifically, by integrating ML techniques with multi-omics data, healthcare
providers can perform the following:
J. Pers. Med. 2024, 14, 931
 Identify biomarkers (e.g., proteins, RNAs, and metabolites) and genetic risk factors22 of 27
that predict the onset of age-related diseases (e.g., Alzheimer’s or CVD) well before
clinical symptoms manifest.
• Explorehow
Explore howchanges
changesin in metabolites
metabolites andandgutgut microbiota
microbiota influence
influence agingaging andrisk
and the the risk
ofofdiseases,
diseases,potentially
potentially leading
leading to to dietary
dietary recommendations,
recommendations, supplementation,
supplementation, or or
probiotictreatments
probiotic treatments toto maintain
maintain health
health andand longevity.
longevity.
• Develop individualized
Develop individualized treatment
treatment plans
plans based
basedonon a person’s
a person’somics profile
omics to maximize
profile to
efficacy and
maximize minimize
efficacy side effects,
and minimize particularly
side effects, for complex
particularly diseases
for complex like cancer
diseases like or
cancer or neurodegenerative
neurodegenerative diseases. diseases.
• Develop
Develop novel treatments that target the
novel treatments that the specific
specific molecular
molecularmechanisms
mechanismsdrivingdrivingaging
aging in each patient, promoting healthy aging and extending life expectancy.
in each patient, promoting healthy aging and extending life expectancy. For example, For
example, a patient with significant telomere attrition might benefit from
a patient with significant telomere attrition might benefit from telomerase activators,telomerase
activators,
telomere telomere protective
protective compounds,compounds, or editing
or editing of theofgenes
the genes affecting
affecting telomere
telomere length
length with CRISPR-CAS9.
with CRISPR-CAS9. AnotherAnother patient
patient withwith mitochondrial
mitochondrial dysfunction
dysfunction could
could receive
receive mitochondrial-targeted
mitochondrial-targeted antioxidants
antioxidants and PGC-1α
and PGC-1α activators.
activators.
ByByleveraging
leveraging the power of
the power of multi-omics
multi-omics data
dataand
andML, ML,wewe
cancan develop
develop precise
precise individ-
individualized treatments that address the unique needs of each aging individual,
ualized treatments that address the unique needs of each aging individual, paving paving
the way
the
forway for a where
a future future healthy
where healthy aging
aging is is achievable
achievable for all.
for all.
The
The future directions arising from the integration of MLofwith
future directions arising from the integration MLmulti-omics
with multi-omics
technologies
technologies are illustrated
are illustrated in Figure 2. in Figure 2.

Futuredirections
Figure2.2.Future
Figure directions
in in
thethe integration
integration of machine
of machine learning
learning withwith multi-omics
multi-omics technologies.
technologies.

6.4.Collaborative
6.4. CollaborativeEfforts
Efforts
in in Data
Data Sharing
Sharing andand Open
Open Science
Science
Collaborativeefforts
Collaborative efforts
in in data
data sharing
sharing andand open
open science
science are are essential
essential for advancing
for advancing
aging research. Sharing data and resources can facilitate the replication and
aging research. Sharing data and resources can facilitate the replication and validation validation
of of
findings, improve the robustness of results, and accelerate the translation of research into into
findings, improve the robustness of results, and accelerate the translation of research
clinicalapplications.
clinical applications.

6.4.1. Data Sharing Platforms

Developing data-sharing platforms that enable researchers to share multi-omics data
and analytical tools can promote collaboration and reproducibility. These platforms should
ensure data privacy and security, provide standardized protocols for data sharing, and
support interoperability between different datasets [104].

6.4.2. Open Science Initiatives

Open science initiatives that promote transparency, accessibility, and reproducibility
in research can enhance the credibility and impact of aging research. These initiatives
include open access to publications, open data repositories, and collaborative research
networks. Supporting open science can facilitate the translation of research findings into
clinical applications and promote the development of personalized interventions for aging
populations [105].
J. Pers. Med. 2024, 14, 931 23 of 27

7. Conclusions
The integration of multi-omics data with ML offers a powerful approach to delve into
a better understanding of the molecular hallmarks of aging. Advances in these technologies
can have a significant impact on aging research. This approach has the potential to identify
critical biomarkers and therapeutic targets, paving the way for personalized interventions
aimed at extending both health span and lifespan. The future of aging research lies in
the seamless integration of these technologies, supported by collaborative efforts in data
sharing and open science. Such advancements will be essential in overcoming current
challenges and translating research findings into clinical practice. Ultimately, by leveraging
the power of multi-omics data and ML, we can move closer to a future where healthy aging
is achievable for all.

Author Contributions: Conceptualization, N.T. and V.S.V.; Methodology, G.F.; Validation, L.T.;
Writing—Original Draft Preparation, N.T., G.F., L.T. and E.P.; Writing—Review and Editing, C.H.,
G.V., V.S.V. and M.N.; Visualization, N.T.; Supervision, V.S.V. and M.N.; Project Administration, M.N.
All authors have read and agreed to the published version of the manuscript.
Funding: This research received no external funding.
Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable.
Data Availability Statement: No new data were created or analyzed in this study. Data sharing is
not applicable to this article.
Conflicts of Interest: The authors declare no conflicts of interest.

References
1. Li, Z.; Zhang, Z.; Ren, Y.; Wang, Y.; Fang, J.; Yue, H.; Ma, S.; Guan, F. Aging and Age-Related Diseases: From Mechanisms to
Therapeutic Strategies. Biogerontology 2021, 22, 165–187. [CrossRef]
2. López-Otín, C.; Blasco, M.A.; Partridge, L.; Serrano, M.; Kroemer, G. Hallmarks of Aging: An Expanding Universe. Cell 2023, 186,
243–278. [CrossRef]
3. Jay, F.B. An Overview of Deep Generative Models in Functional and Evolutionary Genomics. Annu. Rev. Biomed. Data Sci. 2023, 6,
173–189.
4. Evans, M.J.; Smith, R.L. Epigenetic Mechanisms in Aging. Trends Cell Biol. 2021, 31, 347–358.
5. Chambers, D.C.; Carew, A.M.; Lukowski, S.W.; Powell, J.E. Transcriptomics and Single-Cell RNA-Sequencing. Respirology 2019,
24, 29–36. [CrossRef] [PubMed]
6. Moaddel, R.; Ubaida-Mohien, C.; Tanaka, T.; Lyashkov, A.; Basisty, N.; Schilling, B.; Semba, R.D.; Franceschi, C.; Gorospe, M.;
Ferrucci, L. Proteomics in Aging Research: A Roadmap to Clinical, Translational Research. Aging Cell 2021, 20, e13325. [CrossRef]
7. Fang, W.; Chen, S.; Jin, X.; Liu, S.; Cao, X.; Liu, B. Metabolomics in Aging Research: Aging Markers from Organs. Front. Cell Dev.
Biol. 2023, 11, 1071–1078. [CrossRef] [PubMed]
8. Picard, M.; Scott-Boyer, M.-P.; Bodein, A.; Périn, O.; Droit, A. Integration Strategies of Multi-Omics Data for Machine Learning
Analysis. Comput. Struct. Biotechnol. J. 2021, 19, 3735–3746. [CrossRef] [PubMed]
9. Odden, M.C.; Melzer, D. Machine Learning in Aging Research. J. Gerontol. A Biol. Sci. Med. Sci. 2019, 74, 1901–1902. [CrossRef]
10. Nelson, B.C.; Dizdaroglu, M. Implications of DNA Damage and DNA Repair on Human Diseases. Mutagenesis 2020, 35, 1–3.
[CrossRef]
11. Lima, T.; Li, T.Y.; Mottis, A.; Auwerx, J. Pleiotropic Effects of Mitochondria in Aging. Nat. Aging 2022, 2, 199–213. [CrossRef]
[PubMed]
12. Nikfarjam, S.; Singh, K.K. DNA Damage Response Signaling: A Common Link Between Cancer and Cardiovascular Diseases.
Cancer Med. 2023, 12, 4380–4404. [CrossRef]
13. López-Gil, L.; Pascual-Ahuir, A.; Proft, M. Genomic Instability and Epigenetic Changes during Aging. Int. J. Mol. Sci. 2023, 24,
14279. [CrossRef]
14. Xu, Z.; Verma, A.; Naveed, U.; Bakhoum, S.F.; Khosravi, P.; Elemento, O. Deep learning predicts chromosomal instability from
histopathology images. iScience 2021, 24, 102394. [CrossRef] [PubMed]
15. Armanios, M. The Role of Telomeres in Human Disease. Annu. Rev. Genom. Hum. Genet. 2022, 23, 363–381. [CrossRef] [PubMed]
16. Sohn, E.J.; Goralsky, J.A.; Shay, J.W.; Min, J. The Molecular Mechanisms and Therapeutic Prospects of Alternative Lengthening of
Telomeres (ALT). Cancers 2023, 15, 1945. [CrossRef] [PubMed]
J. Pers. Med. 2024, 14, 931 24 of 27

17. Zhang, H.; Kong, W.; Xie, Y.; Zhao, X.; Luo, D.; Chen, S.; Pan, Z. Telomere-related genes as potential biomarkers to predict
endometriosis and immune response: Development of a machine learning-based risk model. Front. Med. 2023, 10, 1132676.
[CrossRef]
18. Zhang, J.; Wang, S.; Liu, B. New Insights into the Genetics and Epigenetics of Aging Plasticity. Genes 2023, 14, 329. [CrossRef]
19. Zhou, J.; Chen, Q.; Braun, P.R.; Mandell, K.A.P.; Jaffe, A.E.; Tan, H.Y.; Hyde, T.M.; Kleinman, J.E.; Potash, J.B.; Shinozaki, G.; et al.
Deep Learning Predicts DNA Methylation Regulatory Variants in the Human Brain and Elucidates the Genetics of Psychiatric
Disorders. Proc. Natl. Acad. Sci. USA 2022, 119, e2206069119. [CrossRef]
20. Higgins-Chen, A.T.; Thrush, K.L.; Wang, Y.; Minteer, C. A Computational Solution for Bolstering Reliability of Epigenetic Clocks:
Implications for Clinical Trials and Longitudinal Tracking. Nat. Aging 2022, 2, 644–661. [CrossRef]
21. Klaips, C.L.; Jayaraj, G.G.; Hartl, F.U. Pathways of Cellular Proteostasis in Aging and Disease. J. Cell Biol. 2018, 217, 51–63.
[CrossRef]
22. Chen, X.Q.; Shen, T.; Fang, S.J.; Sun, X.M.; Li, G.Y.; Li, Y.F. Protein Homeostasis in Aging and Cancer. Front. Cell Dev. Biol. 2023,
11, 1143532. [CrossRef]
23. Thompson, M.A.; De-Souza, E.A. A Year at the Forefront of Proteostasis and Aging. Biol. Open 2023, 12, bio059750. [CrossRef]
24. Eshari, F.; Momeni, F.; Faraj Nezhadi, A.; Shemehsavar, S.; Habibi-Rezaei, M. Prediction of protein aggregation propensity
employing SqFt-based logistic regression model. Int. J. Biol. Macromol. 2023, 249, 126036. [CrossRef] [PubMed]
25. Lu, G.; Wang, Y.; Shi, Y.; Zhang, Z.; Huang, C.; He, W.; Wang, C.; Shen, H.-M. Autophagy in Health and Disease: From Molecular
Mechanisms to Therapeutic Target. MedComm 2022, 3, e150. [CrossRef] [PubMed]
26. Gómez-Virgilio, L.; Silva-Lucero, M.-D.-C.; Flores-Morelos, D.-S.; Gallardo-Nieto, J.; Lopez-Toledo, G.; Abarca-Fernandez, A.-M.;
Zacapala-Gómez, A.-E.; Luna-Muñoz, J.; Montiel-Sosa, F.; Soto-Rojas, L.O.; et al. Autophagy: A Key Regulator of Homeostasis
and Disease: An Overview of Molecular Mechanisms and Modulators. Cells 2022, 11, 2262. [CrossRef] [PubMed]
27. Dong, G.; Gao, H.; Chen, Y.; Yang, H. Machine learning and bioinformatics analysis to identify autophagy-related biomarkers in
peripheral blood for rheumatoid arthritis. Front. Genet. 2023, 14, 1238407. [CrossRef]
28. Johnson, S.C. Nutrient Sensing, Signaling and Ageing: The Role of IGF-1 and mTOR in Ageing and Age-Related Disease. Subcell
Biochem. 2018, 90, 49–97.
29. Fu, W.; Wu, G. Targeting mTOR for Anti-Aging and Anti-Cancer Therapy. Molecules 2023, 28, 3157. [CrossRef]
30. Drewe, J.; Küsters, E.; Hammann, F.; Kreuter, M.; Boss, P.; Schöning, V. Modeling Structure–Activity Relationship of AMPK
Activation. Molecules 2021, 26, 6508. [CrossRef]
31. Videla, L.A.; Marimán, A.; Ramos, B.; José Silva, M.; Del Campo, A. Standpoints in Mitochondrial Dysfunction: Underlying
Mechanisms in Search of Therapeutic Strategies. Mitochondrion 2022, 63, 9–22. [CrossRef] [PubMed]
32. Amorim, J.A.; Coppotelli, G.; Rolo, A.P.; Palmeira, C.M.; Ross, J.M.; Sinclair, D.A. Mitochondrial and Metabolic Dysfunction in
Ageing and Age-Related Diseases. Nat. Rev. Endocrinol. 2022, 18, 243–258. [CrossRef] [PubMed]
33. Qin, H.; Abulaiti, A.; Maimaiti, A.; Abulaiti, Z.; Fan, G.; Aili, Y.; Ji, W.; Wang, Z.; Wang, Y. Integrated machine learning survival
framework develops a prognostic model based on inter-crosstalk definition of mitochondrial function and cell death patterns in a
large multicenter cohort for lower-grade glioma. J. Transl. Med. 2023, 21, 588. [CrossRef] [PubMed]
34. Kuehnemann, C.; Wiley, C.D. Senescent Cells at the Crossroads of Aging, Disease, and Tissue Homeostasis. Aging Cell 2024, 23,
e13988. [CrossRef] [PubMed]
35. Cuollo, L.; Antonangeli, F.; Santoni, A.; Soriani, A. The Senescence-Associated Secretory Phenotype (SASP) in the Challenging
Future of Cancer Therapy and Age-Related Diseases. Biology 2020, 9, 485. [CrossRef] [PubMed]
36. Tuttle, C.S.L.; Waaijer, M.E.C.; Slee-Valentijn, M.S.; Stijnen, T.; Westendorp, R.; Maier, A.B. Cellular Senescence and Chronological
Age in Various Human Tissues: A Systematic Review and Meta-Analysis. Aging Cell 2021, 20, e13083. [CrossRef]
37. Mi, L.; Hu, J.; Li, N.; Gao, J.; Huo, R.; Peng, X.; Zhang, N.; Liu, Y.; Zhao, H.; Liu, R.; et al. The Mechanism of Stem Cell Aging.
Stem Cell Rev. Rep. 2022, 18, 1281–1293. [CrossRef]
38. Weng, Z.; Wang, Y.; Ouchi, T.; Liu, H.; Qiao, X.; Wu, C.; Zhao, Z.; Li, L.; Li, B. Mesenchymal Stem/Stromal Cell Senescence:
Hallmarks, Mechanisms, and Combating Strategies. Stem Cells Transl. Med. 2022, 11, 356–371. [CrossRef]
39. Barardo, D.G.; Newby, D.; Thornton, D.; Ghafourian, T.; de Magalhães, J.P.; Freitas, A.A. Machine Learning for Predicting
Lifespan-Extending Chemical Compounds. Aging 2022, 14, 2163–2189. [CrossRef]
40. Liu, Z.; Liang, Q.; Ren, Y.; Guo, C.; Ge, X.; Wang, L.; Cheng, Q.; Luo, P.; Zhang, Y.; Han, X. Immunosenescence: Molecular
Mechanisms and Diseases. Signal Transduct. Target. Ther. 2023, 8, 200. [CrossRef]
41. Baechle, J.J.; Chen, N.; Makhijani, P.; Winer, S.; Furman, D.; Winer, D.A. Chronic Inflammation and the Hallmarks of Aging. Mol.
Metab. 2023, 74, 101755. [CrossRef] [PubMed]
42. Wang, Y.; Wang, R.; Zhang, S.; Song, S. iTALK: An R Package to Characterize and Illustrate Intercellular Communication. bioRxiv
2021, 507871. [CrossRef]
43. Dugan, B.; Conway, J.; Duggal, N.A. Inflammaging as a Target for Healthy Ageing. Age Ageing 2023, 52, afac328. [CrossRef]
44. Libby, P. Inflammatory Mechanisms: The Molecular Basis of Inflammation and Disease. Nature 2007, 420, 868–874. [CrossRef]
[PubMed]
45. Bobrov, E.; Georgievskaya, A.; Kiselev, K.; Sevastopolsky, A.; Zhavoronkov, A.; Gurov, S.; Rudakov, K.; Del Pilar Bonilla Tobar, M.;
Jaspers, S.; Clemann, S. PhotoAgeClock: Deep Learning Algorithms for Development of Non-Invasive Visual Biomarkers of
Aging. Aging 2022, 14, 1500–1519. [CrossRef] [PubMed]
J. Pers. Med. 2024, 14, 931 25 of 27

46. Salazar, J.; Durán, P.; Díaz, M.P.; Chacín, M.; Santeliz, R.; Mengual, E.; Gutiérrez, E.; León, X.; Díaz, A.; Bernal, M.; et al. Exploring
the Relationship Between the Gut Microbiota and Ageing: A Possible Age Modulator. Int. J. Environ. Res. Public Health 2023, 20,
5845. [CrossRef]
47. Wilmanski, T.; Diener, C.; Rappaport, N.; Patwardhan, S.; Wiedrick, J.; Lapidus, J.; Earls, J.C.; Zimmer, A.; Glusman, G.; Robinson,
M.; et al. Gut Microbiome Pattern Reflects Healthy Ageing and Predicts Survival in Humans. Nat. Metab. 2021, 3, 274–286.
[CrossRef]
48. Shi, Y.; Zhang, H.; Huang, S.; Yin, L.; Wang, F.; Luo, P.; Huang, H. Epigenetic Regulation in Cardiovascular Disease: Mechanisms
and Advances in Clinical Trials. Signal Transduct. Target. Ther. 2022, 7, 200. [CrossRef]
49. Bhole, R.P.; Chikhale, R.V.; Rathi, K.M. Current Biomarkers and Treatment Strategies in Alzheimer Disease: An Overview and
Future Perspectives. IBRO Neurosci. Rep. 2023, 16, 8–42. [CrossRef]
50. Veneziani, I.; Marra, A.; Formica, C.; Grimaldi, A.; Marino, S.; Quartarone, A.; Maresca, G. Applications of Artificial Intelligence
in the Neuropsychological Assessment of Dementia: A Systematic Review. J. Pers. Med. 2024, 14, 113. [CrossRef]
51. Formica, C.; Bonanno, L.; Giambò, F.M.; Maresca, G.; Latella, D.; Marra, A.; Cucinotta, F.; Bonanno, C.; Lombardo, M.; Tomarchio,
O.; et al. Paving the Way for Predicting the Progression of Cognitive Decline: The Potential Role of Machine Learning Algorithms
in the Clinical Management of Neurodegenerative Disorders. J. Pers. Med. 2023, 13, 1386. [CrossRef] [PubMed]
52. Irtyuga, O.; Babakekhyan, M.; Kostareva, A.; Uspensky, V.; Gordeev, M.; Faggian, G.; Malashicheva, A.; Metsker, O.; Shlyakhto,
E.; Kopanitsa, G. Analysis of Prevalence and Clinical Features of Aortic Stenosis in Patients with and Without Bicuspid Aortic
Valve Using Machine Learning Methods. J. Pers. Med. 2023, 13, 1588. [CrossRef]
53. Paluch-Shimon, S.; Cardoso, F. PARP Inhibitors Coming of Age. Nat. Rev. Clin. Oncol. 2021, 18, 69–70. [CrossRef] [PubMed]
54. Sharma, A.; Chabloz, S.; Lapides, R.A.; Roider, E.; Ewald, C.Y. Potential Synergistic Supplementation of NAD+ Promoting
Compounds as a Strategy for Increasing Healthspan. Nutrients 2023, 15, 445. [CrossRef]
55. Derevyanko, A.; Skowronska, A.; Skowronski, M.T.; Kordowitzki, P. The Interplay Between Telomeres, Mitochondria, and
Chronic Stress Exposure in the Aging Egg. Cells 2022, 11, 2612. [CrossRef] [PubMed]
56. Cao, Q.; Tian, Y.; Deng, Z.; Yang, F.; Chen, E. Epigenetic Alteration in Colorectal Cancer: Potential Diagnostic and Prognostic
Implications. Int. J. Mol. Sci. 2024, 25, 3358. [CrossRef] [PubMed]
57. Wang, K.; Liu, H.; Hu, Q.; Wang, L.; Liu, J.; Zheng, Z.; Zhang, W.; Ren, J.; Zhu, F.; Liu, G.-H. Epigenetic Regulation of Aging:
Implications for Interventions of Aging and Diseases. Signal Transduct. Target. Ther. 2022, 7, 374. [CrossRef]
58. Türker, F.; Cook, E.K.; Margolis, S.S. The Proteasome and Its Role in the Nervous System. Cell Chem. Biol. 2021, 28, 903–917.
[CrossRef]
59. Kaushik, S.; Tasset, I.; Arias, E.; Pampliega, O.; Wong, E.; Martinez-Vicente, M.; Cuervo, A.M. Autophagy and the Hallmarks of
Aging. Ageing Res. Rev. 2021, 72, 101468. [CrossRef]
60. Sharp, Z.D.; Strong, R. Rapamycin, the Only Drug That Has Been Consistently Demonstrated to Increase Mammalian Longevity:
An Update. Exp. Gerontol. 2023, 176, 112166. [CrossRef]
61. Zhu, X.; Shen, W.; Liu, Z.; Sheng, S.; Xiong, W.; He, R.; Zhang, X.; Ma, L.; Ju, Z. Effect of Metformin on Cardiac Metabolism and
Longevity in Aged Female Mice. Front. Cell Dev. Biol. 2021, 8, 626011. [CrossRef] [PubMed]
62. Iside, C.; Scafuro, M.; Nebbioso, A.; Altucci, L. SIRT1 Activation by Natural Phytochemicals: An Overview. Front. Pharmacol.
2020, 11, 1225. [CrossRef] [PubMed]
63. Norat, P.; Soldozy, S.; Sokolowski, J.D.; Gorick, C.M.; Kumar, J.S.; Chae, Y.; Yağmurlu, K.; Prada, F.; Walker, M.; Levitt, M.R.; et al.
Mitochondrial Dysfunction in Neurological Disorders: Exploring Mitochondrial Transplantation. NPJ Regen. Med. 2020, 5, 22.
[CrossRef]
64. Lu, Y.; Fujioka, H.; Wang, W.; Zhu, X. Bezafibrate Confers Neuroprotection in the 5xFAD Mouse Model of Alzheimer’s Disease.
Biochim. Biophys. Acta Mol. Basis Dis. 2023, 1869, 166841. [CrossRef] [PubMed]
65. Kirkland, J.L.; Tchkonia, T. Senolytic Drugs: From Discovery to Translation. J. Intern. Med. 2020, 288, 518–536. [CrossRef]
[PubMed]
66. Xu, M.; Tchkonia, T.; Kirkland, J.L.; Nakahara, T. Senolytics Improve Physical Function and Increase Lifespan in Old Age. Nat.
Med. 2018, 21, 857–862. [CrossRef]
67. O’Loghlen, A. The Potential of Aging Rejuvenation. Cell Cycle 2022, 21, 111–116. [CrossRef]
68. Jiang, T.; Gradus, J.L.; Rosellini, A.J. Supervised Machine Learning: A Brief Primer. Behav. Ther. 2020, 51, 675–687. [CrossRef]
69. Eckhardt, C.M.; Madjarova, S.J.; Williams, R.J.; Ollivier, M.; Karlsson, J.; Pareek, A.; Nwachukwu, B.U. Unsupervised Machine
Learning Methods and Emerging Applications in Healthcare. Knee Surg. Sports Traumatol. Arthrosc. 2023, 31, 376–381. [CrossRef]
70. Al-Hamadani, M.N.A.; Fadhel, M.A.; Alzubaidi, L.; Balazs, H. Reinforcement Learning Algorithms and Applications in Healthcare
and Robotics: A Comprehensive and Systematic Review. Sensors 2024, 24, 2461. [CrossRef]
71. LeCun, Y.; Bengio, Y.; Hinton, G. Deep Learning. Nature 2015, 521, 436–444. [CrossRef] [PubMed]
72. Subramanian, I.; Verma, S.; Kumar, S.; Jere, A.; Anamika, K. Multi-Omics Data Integration, Interpretation, and Its Application.
Bioinform. Biol. Insights 2020, 14, 1177932219899051. [CrossRef] [PubMed]
73. Athieniti, E.; Spyrou, G.M. A Guide to Multi-Omics Data Collection and Integration for Translational Medicine. Comput. Struct.
Biotechnol. J. 2022, 20, 134–149. [CrossRef]
74. Zitnik, M.; Agrawal, M.; Leskovec, J. Modeling Polypharmacy Side Effects with Graph Convolutional Networks. Bioinformatics
2018, 34, i457–i466. [CrossRef] [PubMed]
J. Pers. Med. 2024, 14, 931 26 of 27

75. Theodoris, C.V.; Xiao, L.; Chopra, A.; Chaffin, M.D.; Al Sayed, Z.R.; Hill, M.C.; Mantineo, H.; Brydon, E.M.; Zeng, Z.; Liu, X.S.;
et al. Transfer Learning Enables Predictions in Network Biology. Nature 2023, 618, 616–624. [CrossRef] [PubMed]
76. Wu, W.-T.; Li, Y.-J.; Feng, A.-Z.; Li, L.; Huang, T.; Xu, A.-D.; Lyu, J. Data Mining in Clinical Big Data: The Frequently Used
Databases, Steps, and Methodological Models. Mil. Med. Res. 2021, 8, 44. [CrossRef]
77. Ketkar, Y.; Gawade, S. A Decision Support System for Selecting the Most Suitable Machine Learning in Healthcare Using User
Parameters and Requirements. Healthc. Anal. 2022, 2, 100117. [CrossRef]
78. Huang, S.; Cai, N.; Pacheco, P.P.; Narrandes, S.; Wang, Y.; Xu, W. Applications of Support Vector Machine (SVM) Learning in
Cancer Genomics. Cancer Genom. Proteom. 2018, 15, 41–51.
79. Quist, J.; Taylor, L.; Staaf, J.; Grigoriadis, A. Random Forest Modelling of High-Dimensional Mixed-Type Data for Breast Cancer
Classification. Cancers 2021, 13, 991. [CrossRef]
80. Navarro, S.; Ventura, S. Computational Methods to Predict Protein Aggregation. Curr. Opin. Struct. Biol. 2022, 73, 102343.
[CrossRef]
81. Whittemore, K.; Derevyanko, A.; Martinez, P.; Serrano, R.; Pumarola, M.; Bosch, F.; Blasco, M.A. Telomerase Gene Therapy
Ameliorates the Effects of Neurodegeneration Associated with Short Telomeres in Mice. Aging 2019, 11, 2916–2948. [CrossRef]
[PubMed]
82. Squarzoni, S.; Schena, E.; Sabatelli, P.; Mattioli, E.; Capanni, C.; Cenni, V.; D’Apice, M.R.; Andrenacci, D.; Sarli, G.; Pellegrino, V.;
et al. Interleukin-6 Neutralization Ameliorates Symptoms in Prematurely Aged Mice. Aging Cell 2021, 20, e13285. [CrossRef]
[PubMed]
83. Liu, J.; Wang, Y.J. Rejuvenating the Immune System: Insights for Anti-Neurodegeneration Strategies. Neurosci. Bull. 2022, 38,
107–109. [CrossRef] [PubMed] [PubMed Central]
84. Zhang, L.; Zhao, J.; Mu, X.; McGowan, S.J.; Angelini, L.; O’Kelly, R.D.; Yousefzadeh, M.J.; Sakamoto, A.; Aversa, Z.; LeBrasseur,
N.K.; et al. Novel small molecule inhibition of IKK/NF-κB activation reduces markers of senescence and improves healthspan in
mouse models of aging. Aging Cell 2021, 20, e13486. [CrossRef] [PubMed]
85. Dagdeviren, S.; Jung, D.Y.; Friedline, R.H.; Noh, H.L.; Kim, J.H.; Patel, P.R.; Tsitsilianos, N.; Inashima, K.; Tran, D.A.; Hu, X.; et al.
IL-10 prevents aging-associated inflammation and insulin resistance in skeletal muscle. FASEB J. 2017, 31, 701–710. [CrossRef]
[PubMed]
86. Parker, A.; Romano, S.; Ansorge, R.; Aboelnour, A.; Le Gall, G.; Savva, G.M.; Pontifex, M.G.; Telatin, A.; Baker, D.; Jones, E.; et al.
Fecal microbiota transfer between young and aged mice reverses hallmarks of the aging gut, eye, and brain. Microbiome 2022, 10,
68. [CrossRef]
87. Shapiro, S.C. Biomarkers in Rheumatoid Arthritis. Cureus 2021, 13, e15063. [CrossRef]
88. Hsu, L.-C.; Li, C.-J.; Lin, L.-T.; Pan, L.-F.; Wen, Z.-H.; Sheu, J.J.-C.; Tsui, K.-H. Multi-Omics Reveals the Role of Osteopon-
tin/Secreted Phosphoprotein 1 in Regulating Ovarian Aging. J. Pers. Med. 2024, 14, 78. [CrossRef]
89. Carrillo-Perez, F.; Morales, J.C.; Castillo-Secilla, D.; Gevaert, O.; Rojas, I.; Herrera, L.J. Machine-Learning-Based Late Fusion on
Multi-Omics and Multi-Scale Data for Non-Small-Cell Lung Cancer Diagnosis. J. Pers. Med. 2022, 12, 601. [CrossRef]
90. Sureshkumar, V.; Prasad, R.S.N.; Balasubramaniam, S.; Jagannathan, D.; Daniel, J.; Dhanasekaran, S. Breast Cancer Detection and
Analytics Using Hybrid CNN and Extreme Learning Machine. J. Pers. Med. 2024, 14, 792. [CrossRef]
91. Patrascanu, O.S.; Tutunaru, D.; Musat, C.L.; Dragostin, O.M.; Fulga, A.; Nechita, L.; Ciubara, A.B.; Piraianu, A.I.; Stamate, E.;
Poalelungi, D.G.; et al. Future Horizons: The Potential Role of Artificial Intelligence in Cardiology. J. Pers. Med. 2024, 14, 656.
[CrossRef] [PubMed]
92. Ming, C.; Lee, G.J.W.; Teo, Y.H.; Teo, Y.N.; Toh, E.M.S.; Li, T.Y.W.; Guo, C.Y.; Ding, J.; Zhou, X.; Teoh, H.L.; et al. Machine Learning
Modeling to Predict Atrial Fibrillation Detection in Embolic Stroke of Undetermined Source Patients. J. Pers. Med. 2024, 14, 534.
[CrossRef]
93. Elvas, L.B.; Nunes, M.; Ferreira, J.C.; Dias, M.S.; Rosário, L.B. AI-Driven Decision Support for Early Detection of Cardiac Events:
Unveiling Patterns and Predicting Myocardial Ischemia. J. Pers. Med. 2023, 13, 1421. [CrossRef] [PubMed]
94. Tapper, W.; Carneiro, G.; Mikropoulos, C.; Thomas, S.A.; Evans, P.M.; Boussios, S. The Application of Radiomics and AI to
Molecular Imaging for Prostate Cancer. J. Pers. Med. 2024, 14, 287. [CrossRef] [PubMed]
95. Martínez-García, M.; Hernández-Lemus, E. Data Integration Challenges for Machine Learning in Precision Medicine. Front. Med.
2022, 8, 784455. [CrossRef]
96. Bernardi, F.A.; Alves, D.; Crepaldi, N.; Yamada, D.B.; Lima, V.C.; Rijo, R. Data Quality in Health Research: Integrative Literature
Review. J. Med. Internet Res. 2023, 25, e41446. [CrossRef]
97. Vikas, H.; Kini, A.; Sharma, N.; Gowda, N.R.; Gupta, A. How informed is the informed consent? J. Fam. Med. Prim. Care 2021, 10,
2299–2303. [CrossRef]
98. Yadav, N.; Pandey, S.; Gupta, A.; Dudani, P.; Gupta, S.; Rangarajan, K. Data Privacy in Healthcare: In the Era of Artificial
Intelligence. Indian Dermatol. Online J. 2023, 14, 788–792. [CrossRef]
99. Wiltshire, D.; Alvanides, S. Ensuring the Ethical Use of Big Data: Lessons from Secure Data Access. Heliyon 2022, 8, e08981.
[CrossRef]
100. Washington, V.; Franklin, J.B.; Huang, E.S.; Mega, J.L.; Abernethy, A.P. Diversity, Equity, and Inclusion in Clinical Research: A
Path Toward Precision Health for Everyone. Clin. Pharmacol. Ther. 2023, 113, 575–584. [CrossRef]
J. Pers. Med. 2024, 14, 931 27 of 27

101. Stein, C.M.; Weiskirchen, R.; Damm, F.; Strzelecka, P.M. Single-Cell Omics: Overview, Analysis, and Application in Biomedical
Science. J. Cell. Biochem. 2021, 122, 1571–1578. [CrossRef]
102. Bressan, D.; Battistoni, G.; Hannon, G.J. The Dawn of Spatial Omics. Science 2023, 381, eabq4964. [CrossRef] [PubMed]
103. Allen, B. The Promise of Explainable AI in Digital Health for Precision Medicine: A Systematic Review. J. Pers. Med. 2024, 14, 277.
[CrossRef] [PubMed]
104. Devriendt, T.; Borry, P.; Shabani, M. Factors That Influence Data Sharing Through Data Sharing Platforms: A Qualitative Study
on the Views and Experiences of Cohort Holders and Platform Developers. PLoS ONE 2021, 16, e0254202. [CrossRef] [PubMed]
105. Isaacowitz, D.M.; Lind, M. Open Science Is for Aging Research, Too. Innov. Aging 2019, 3, igz028. [CrossRef]

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual
author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to
people or property resulting from any ideas, methods, instructions or products referred to in the content.