Intronic polyadenylation (IPA) is a key mechanism driving transcriptome diversity, yet its detection and functional characterization remain challenging due to complex splicing patterns and complexity of intronic regions. Here, we introduce IPAseek, a dynamic programming-based computational framework that leverages the Pruned Exact Linear Time (PELT) algorithm and Changepoints Over a Range of PenaltieS (CROPS) to enable de novo identification of IPA events from bulk RNA-seq data. IPAseek robustly detects both composite and skipped IPA isoforms. Applying IPAseek to bulk RNA-seq of hematopoietic cell types, reveals lineage and stage-specific IPA signatures, with lymphoid cells exhibiting higher IPA site usage than myeloid cells. Temporal profiling during megakaryocyte differentiation uncovers dynamic, gene-specific IPA regulation linked to functional pathways including peroxisomal metabolism and autophagy which are known to play a crucial role in megakaryocytic differentiation, impacting the development and maturation of megakaryocytes. Further, integrative analysis demonstrates that IPA site usage is associated with lower DNA methylation within introns, supporting a regulatory axis connecting epigenetic state and IPA. This finding aligns with emerging evidence that DNA methylation modulates alternative polyadenylation via CTCF-mediated chromatin looping. Thus, IPAseek provides a platform to characterize IPA across physiological systems and disease contexts using widely available bulk RNA-seq data. These IPA events can be further integrated with other regulatory datasets to elucidate their interplay and functional significance.
Rashmi et al. (2026) studied this question.