Splam: a deep-learning-based splice site predictor that improves spliced alignments

Kuan-Hao Chao; Alan Mao; Steven L Salzberg; Mihaela Pertea

doi:10.1186/s13059-024-03379-4

Splam: a deep-learning-based splice site predictor that improves spliced alignments

Genome Biol. 2024 Sep 16;25(1):243. doi: 10.1186/s13059-024-03379-4.

Authors

Kuan-Hao Chao^{1

2}, Alan Mao^{3

4

5}, Steven L Salzberg^{3

4

5

6}, Mihaela Pertea^{7

8

9}

Affiliations

¹ Department of Computer Science, Johns Hopkins University, Baltimore, MD, 21218, USA. [email protected].
² Center for Computational Biology, Johns Hopkins University, Baltimore, MD, 21211, USA. [email protected].
³ Department of Computer Science, Johns Hopkins University, Baltimore, MD, 21218, USA.
⁴ Center for Computational Biology, Johns Hopkins University, Baltimore, MD, 21211, USA.
⁵ Department of Biomedical Engineering, Johns Hopkins University, Baltimore, MD, 21218, USA.
⁶ Department of Biostatistics, Johns Hopkins University, Baltimore, MD, 21205, USA.
⁷ Department of Computer Science, Johns Hopkins University, Baltimore, MD, 21218, USA. [email protected].
⁸ Center for Computational Biology, Johns Hopkins University, Baltimore, MD, 21211, USA. [email protected].
⁹ Department of Biomedical Engineering, Johns Hopkins University, Baltimore, MD, 21218, USA. [email protected].

Abstract

The process of splicing messenger RNA to remove introns plays a central role in creating genes and gene variants. We describe Splam, a novel method for predicting splice junctions in DNA using deep residual convolutional neural networks. Unlike previous models, Splam looks at a 400-base-pair window flanking each splice site, reflecting the biological splicing process that relies primarily on signals within this window. Splam also trains on donor and acceptor pairs together, mirroring how the splicing machinery recognizes both ends of each intron. Compared to SpliceAI, Splam is consistently more accurate, achieving 96% accuracy in predicting human splice junctions.

Keywords: Convolutional neural network; Deep learning; Dilated convolution; Residual block; Splice site predictor; Spliced alignment.

MeSH terms

Deep Learning*
Humans
Introns
Neural Networks, Computer
RNA Splice Sites*
RNA Splicing*
Sequence Alignment

Substances

RNA Splice Sites