BIMSA >
Seminar on Bioinformatics
Self-supervised learning on millions of pre-mRNA sequences improves sequence-based RNA splicing prediction
Self-supervised learning on millions of pre-mRNA sequences improves sequence-based RNA splicing prediction
Organizer
Speaker
Time
Thursday, October 26, 2023 10:00 PM - 10:30 PM
Venue
Online
Abstract
RNA splicing, a crucial post-transcriptional process in eukaryotic gene expression, can be predicted using a novel self-supervised model, SpliceBERT. Pre-trained on vertebrate pre-mRNA sequences, SpliceBERT enhances RNA splicing prediction, including variant effects and branch point prediction, across species, emphasizing the potential of pre-trained models for understanding RNA splicing.