Hierarchical clustering-based coarse-to-fine classification framework for microbial protein function prediction
Shengyang Chen, Xinyue Gao, Congmin Zhu, Honglei Liu, Yuqing Yang · 2025
Read Hierarchical clustering-based coarse-to-fine classification framework for microbial protein function prediction by Shengyang Chen, Xinyue Gao, Congmin Zhu, Honglei Liu, Yuqing Yang (2025) 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-12750710.txt.
Description
Shengyang Chen and co-authors build, in a 2025 paper, a coarse-to-fine classifier that predicts what microbial proteins do, with special care for rare functions. Knowing protein jobs drives probiotic discovery and biotechnology, but function labels form deep family trees and most branches are rare. Standard deep learning flattens that tree and fails on the long tail. The authors' network instead climbs the tree level by level, letting parent and child labels inform each other.
For one label system the cascade follows the official hierarchy directly. For the other, it builds a hierarchy from meaning-based clustering of the terms themselves, then attends across levels. Skewed training data is rebalanced on the fly, and rare labels get extra weight in the loss. Across all major categories the system beat both classic sequence matching and rival neural tools.
Catalog record
| CREATOR | Shengyang Chen, Xinyue Gao, Congmin Zhu, Honglei Liu, Yuqing Yang |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 47 KB |
| VERSION | 1.0 |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |