CATaN framework jointly models transcription factor networks and transcriptomes to map complex disease heritability

Researchers have posted a preprint to bioRxiv describing CATaN (Causal variant Analysis via Transcription factor Networks), a framework designed to bridge two streams of genomic analysis that have largely been pursued separately: transcription factor gene regulatory networks (TF-GRNs) and transcriptome-wide association data.

Genome-wide association studies (GWAS) have established that causal variants for complex traits are disproportionately enriched at transcription factor binding sites, suggesting that many disease-associated variants act by disrupting TF activity and, in turn, altering gene expression programmes. However, existing methods typically assess TF-GRNs or transcriptomes independently. CATaN constructs a matrix jointly encoding TF-GRN structure and transcriptomic data, then uses this representation to quantify how much of a trait's heritability can be attributed to specific regulatory programmes.

The authors report that CATaN can identify disease-relevant regulatory programmes across multiple complex diseases, potentially helping to prioritise causal variants and the transcription factors that mediate their effects. The preprint has not yet been peer-reviewed. Methods of this kind are relevant to statistical genetics and computational genomics researchers working on GWAS fine-mapping, gene regulation, and the functional interpretation of non-coding variants — a persistent challenge in translating association signals into mechanistic understanding.