Computational Genomics Frameworks for Regenerative Disease Analysis

Abstract

Regenerative diseases -- conditions characterised by chronic failure of tissue repair, including muscular dystrophies, degenerative joint disease, chronic liver fibrosis, age-related macular degeneration, and spinal cord injury -- share a common genomic architecture of disrupted repair pathways, altered epigenetic regulation of regeneration genes, and somatic mutation accumulation that compounds regenerative failure over time. Computational genomics -- the application of bioinformatics and machine learning to large-scale genomic, transcriptomic, and epigenomic datasets -- provides tools for systematically characterising this genomic landscape and identifying therapeutic targets and patient stratification markers at scale impossible with traditional molecular biology. This paper proposes the Regenerative Disease Computational Genomics (RDCG) framework, a comprehensive bioinformatics methodology for the systematic analysis of the genomic and epigenomic underpinnings of regenerative failure, comprising five analytical modules: a variant effect predictor for regeneration genes (VEPRG) that prioritises disease-causing variants in regeneration pathway genes; an epigenomic dysregulation analyser (EDA) that identifies aberrant chromatin accessibility and DNA methylation patterns at regeneration gene loci; a gene regulatory network reconstructor (GRNR) that infers the transcription factor networks controlling regeneration gene expression in diseased tissues; a polygenic regeneration score calculator (PRSC) that aggregates common variant effects into a genomic predictor of regenerative capacity; and a therapeutic target prioritiser (TTP) that integrates multi- level genomic evidence to rank candidate therapeutic targets. RDCG is applied to four regenerative disease datasets -- Duchenne muscular dystrophy (DMD), age-related macular degeneration (AMD), osteoarthritis (OA), and spinal cord injury (SCI) -- comprising 18,642 individuals with whole-genome sequencing, RNA-seq, and ATAC-seq data. RDCG identifies 284 high-confidence therapeutic target candidates across the four conditions, 48 of which have available drug compounds in DrugBank, representing immediate repurposing opportunities. The polygenic regeneration score achieves AUC = 0.842 (SD = 0.028) for regenerative capacity prediction, outperforming single-variant predictors (AUC = 0.684). The study contributes the RDCG specification, the RegenomeDB genomic database, and a ranked catalogue of 284 computational- genomics-identified therapeutic targets for four regenerative diseases.