Zoe Chervontseva1*, Jan Baumbach1 and Olga Tsoy2
1University of Hamburg
2Vrije Universiteit Amsterdam
zoia.chervontseva [at] uni-hamburg.de
Abstract
Aging affects many molecular traits, but it remains difficult to separate shared regulatory trends from dataset noise and lineage effects. We focus on alternative splicing, a major post-transcriptional regulatory layer that changes with age in many tissues. Our aim is to test how conserved age-related splicing regulation is across vertebrates and whether recurring patterns point to common regulatory mechanisms.
We are assembling a multi-tissue, multi-species collection of bulk RNA-seq datasets, including human GTEx and additional vertebrates such as mouse, macaque, chicken, and naked mole-rat. Because read support for individual splicing events is often limited in typical RNA-seq, we use count-based modeling to estimate age-associated trends while accounting for sampling uncertainty. In human tissues, preliminary results recover broad age-associated splicing changes and suggest that splicing variability across individuals often increases with age, consistent with reduced robustness of splicing control.
We then map candidate orthologous events across species using genome alignments and perform tissue-matched comparisons where possible. A key idea is that cross-species recurrence can serve as a practical filter: events that show similar age trends in multiple species are less likely to reflect dataset-specific artifacts. Early comparative analyses already highlight candidate conserved age-associated events. Initial functional annotation suggests that some of these events overlap protein regions linked to protein-protein interactions.
Ongoing work expands the analysis to additional event types and improves mechanism prioritization by integrating predicted NMD sensitivity, available chromatin features, and splice factor based predictors, with orthogonal support from perturbation datasets. This work in progress aims to produce a comparative atlas of age-associated splicing and a framework for exploring conserved regulatory mechanisms of aging.
Keywords: aging, alternative splicing, mechanistic inference