Answers are grounded in this paper and cite their sources.
A stochastic model demonstrates the probabilistic nature of splicing, influencing transcript diversity in physiology and disease.
The human proteome exhibits substantially greater complexity than expected from the ∼20,000 protein-coding genes, a discrepancy largely attributed to alternative splicing (AS). While AS is generally viewed as tightly regulated by cis-elements and trans-acting factors, the degree to which isoform diversity arises from the intrinsic properties of the splicing process itself remains unresolved. We present a minimal stochastic model of RNA splicing in which splice site recognition follows intrinsic probabilistic rules, and splicing proceeds iteratively through the earliest recognized compatible splice sites until termination. Despite its simplicity, the model reproduces key transcriptome-wide features: isoform distributions dominated by a single major isoform per gene, exon skipping as the most frequent AS event, and remarkable robustness of full-length isoforms to perturbations of splice site recognition weights. Predictions align quantitatively with transcriptome-scale patterns from GTEx, spanning exon classes and tissue-specific splicing profiles, and accurately reproduce canonical splicing outcomes. Furthermore, features such as exon definition emerge spontaneously from the model, without requiring explicit regulatory rules. These findings establish a unifying framework in which regulation acts as a modulatory layer superimposed on a fundamentally stochastic process, providing a quantitative baseline for distinguishing regulated from stochastic contributions to transcript diversity in physiology and disease.
No takes yet. Share an insight, caveat, or question.
Verley et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: