Computational Genomics of Longevity and Healthy Aging

Abstract

Aging is the primary risk factor for most chronic diseases, including heart disease, cancer, and neurodegeneration. Instead of tackling these diseases one by one, a more powerful approach is to target the underlying biological aging process itself. This Masters in Computer Science proposal outlines a computational and systems biology approach to identify and validate novel therapeutic targets for extending healthspan (the period of life spent in good health). The research will focus on integrating multi-omics data (genomics, transcriptomics, proteomics) from diverse species and human cohorts to build a predictive model of aging and identify key molecular pathways that can be targeted for intervention.

Key Research Questions in Geroscience

1. Cross-Species Genomic Signatures of Longevity: What are the common genetic and gene expression patterns shared among long-lived species (e.g., naked mole-rats, bowhead whales) but absent in shorter-lived ones? Can we use comparative genomics to identify novel longevity-associated genes?
2. Predictive "Aging Clocks": Can we build a more accurate molecular biomarker of biological age (an "aging clock") by training a machine learning model on multi-omics data? The goal is to create a model that can predict an individual's healthspan and mortality risk more accurately than chronological age.
3. Causal Inference of Aging Pathways: Using techniques like Mendelian Randomization on large human genetic datasets (like the UK Biobank), can we determine which molecular pathways have a causal effect on aging and age-related diseases, separating them from mere correlations?
4. In-Silico Drug Screening: Can we use the identified causal pathways to computationally screen for existing FDA-approved drugs that could be repurposed as geroprotectors (drugs that slow aging)?

Proposed Masters in Computer Science Research: A Multi-Omic Aging Clock and Drug Discovery Platform

This thesis will focus on building a computational pipeline to discover and prioritize anti-aging interventions.

• Data Integration: Assemble a large, harmonized database of publicly available multi-omics data from different tissues, ages, and species.
• Development of a Multi-Omic Aging Clock: Train a deep learning model (e.g., a variational autoencoder) on this dataset to create a robust aging clock. This clock will be validated on independent human cohort data to ensure its predictive power for health outcomes.
• Causal Network Analysis: Use the aging clock as a quantitative phenotype. Apply Mendelian Randomization and other causal inference methods to identify genes and pathways that have a causal impact on the rate of aging as measured by the clock.
• Computational Drug Repurposing: Create a "connectivity map" that links the gene expression signatures of aging with the signatures of thousands of known drugs. This will be used to identify drugs that have a gene expression profile that is the "opposite" of the aging signature, making them prime candidates for geroprotectors. The top candidates will be proposed for experimental validation in cell cultures or animal models.

Impact for Ghana and Africa

Africa is projected to experience the fastest growth in its elderly population of any continent. This will lead to a massive increase in the burden of age-related chronic diseases. This research is directly relevant to this challenge. While the computational work is universal, the developed aging clock could be fine-tuned and validated on data from Ghanaian populations to understand the specific drivers of aging in an African context. More importantly, by identifying low-cost, existing drugs that can be repurposed to promote healthy aging, this research offers a pragmatic and affordable strategy for Ghana to improve the healthspan of its citizens. It would shift the medical paradigm from treating individual diseases late in life to preventing them by targeting the aging process itself, a strategy that is both more effective and more economically sustainable for a developing healthcare system.