Inferring position-dependent splicing factor activity from intronic sequence conservation
Name
rotellokuri-nrotello-ms-biology-2025-thesis.pdf
Size
5.09 MB
Format
Adobe PDF
Checksum (MD5)
788e00623a2510847fc81be077bc02c2
Author(s)
Rotello-Kuri, Nicholas N.
Advisor(s)
Burge, Christopher B.
Date Issued
September 2025
Publisher
Massachusetts Institute of Technology
Abstract
Pre-mRNA undergoes splicing to remove introns and ligate exons to form mature mRNA. Splicing regulation can lead to modulation of protein activity and gene expression. Along with the core 5’ and 3’ splice sites, splicing regulation is affected by conserved splicing factors. The impact of a splicing factor on exon inclusion depends on the binding position relative to a splice site. Prior work has evaluated the position-dependent impacts of 100+ splicing factors via knockdown RNA-seq and eCLIP to generate activity maps, where positive activity promotes exon inclusion, while negative activity promotes exon skipping. However, relying solely on knockdown and eCLIP to create splicing factor activity maps has limited the ability to compare knockdown-derived activity maps with an orthogonal approach. Furthermore, knockdown may have indirect effects related to cell viability, homolog compensation, and perturbing gene regulatory networks, therefore we developed an alternative approach to construct splicing factor activity maps. To address these limitations, we introduce an orthogonal strategy for constructing splicing factor activity maps that leverages evolutionary conservation. Exonic regions are constrained by their protein-coding potential, while intronic regions are generally more variable. However, intronic sequence conservation near splice sites frequently reflects functional regulatory elements, particularly splicing factor binding motifs. By integrating comparative genomics with mammalian tissue RNA-seq datasets, we systematically analyze how conserved intronic sequences correlate with the positional activity of splicing factors RBFOX, HNRNP K, HNRNP C, PCBP1, and PTBP1. Our approach reveals how evolutionary constraints can be used to infer splicing factor function in a position-dependent manner, offering a robust, orthogonal framework to validate and extend existing activity maps. This conservation-informed model enhances our understanding of the regulatory grammar of splicing and provides a foundation for studying splicing regulation across species and cell types without requiring perturbation of splicing factor activity.
MIT Department
Massachusetts Institute of Technology. Department of Biology
Terms of Use
In Copyright - Educational Use Permitted
Copyright retained by author(s)
Persistent DSpace Link
