Spliceread: improving canonical and non-canonical splice site prediction with residual blocks and synthetic data augmentation
Sahil Thapa, Khushali Samderiya, Rohit Menon, Oluwatosin Oluwadare · 2026
Read Spliceread: improving canonical and non-canonical splice site prediction with residual blocks and synthetic data augmentation by Sahil Thapa, Khushali Samderiya, Rohit Menon, Oluwatosin Oluwadare (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/pmc-13390325.txt.
Description
Sahil Thapa's team presents SpliceRead, a deep-learning model. It predicts both common and rare splice sites in gene strings, where most tools favour frequent usual sites and miss key non-usual variants in living things. The model pairs spare layered blocks with made-up data boosting that generates varied non-usual examples.
Trained on multi-animal strings of four hundred and six hundred bases, SpliceRead beat top models on hit rate, exact hits, recall, and F1 score. Its non-usual miscall rate was much lower. Cross-validation, cross-species tests, and length spread proved strength, and the boosting trick also better rival models.
Catalog record
| CREATOR | Sahil Thapa, Khushali Samderiya, Rohit Menon, Oluwatosin Oluwadare |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 60 KB |
| VERSION | 1.0 |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |