Protein Function Prediction: From Traditional Classifier to Deep Learning
Problem
Predicting the biological function of proteins from sequence and interaction data, a task where traditional classifiers underperform.
Why it matters
Knowing protein functions underpins drug discovery and understanding of disease mechanisms.
What the AI did
DeepFunc, a deep learning framework using derived feature information from protein sequences and protein interaction networks, predicted protein functions more accurately than the prior DeepGO method.
Results
DeepFunc more accurate than DeepGO; combining multiple derived features outperformed single derived features
Impact
Advanced deep-learning-based protein function prediction accuracy over a previously reported method.
Record
| Venue | PROTEOMICS |
|---|---|
| Published | 2019 |
| Work type | article |
| Authors | Zhibin Lv, Chunyan Ao, Quan Zou |
| Citations | 132 (OpenAlex, at import time) |
| References | 11 |
| Model or method | DeepFunc (deep learning framework) |
| Organization | not stated in source |
| Primary source | https://doi.org/10.1002/pmic.201900119 |
| Open access | closed |
| Keywords | deep learning, protein function prediction, DeepFunc, DeepGO, bioinformatics |
Importance rating
real but narrower advance
Protein function prediction is a foundational bioinformatics task with relevance to drug discovery and disease understanding, but the abstract only shows an accuracy improvement over a prior method (DeepGO); no scale stated.