Contrastive Representation Learning for 3D Protein Structures
This addresses data scarcity in structural bioinformatics for researchers and practitioners, though it is incremental as it applies existing contrastive learning to a new domain.
The paper tackles the challenge of limited annotated 3D protein structures by introducing an unsupervised contrastive learning framework to learn representations from the Protein Data Bank, achieving new state-of-the-art results in tasks like protein function prediction and fold classification.
Learning from 3D protein structures has gained wide interest in protein modeling and structural bioinformatics. Unfortunately, the number of available structures is orders of magnitude lower than the training data sizes commonly used in computer vision and machine learning. Moreover, this number is reduced even further, when only annotated protein structures can be considered, making the training of existing models difficult and prone to over-fitting. To address this challenge, we introduce a new representation learning framework for 3D protein structures. Our framework uses unsupervised contrastive learning to learn meaningful representations of protein structures, making use of proteins from the Protein Data Bank. We show, how these representations can be used to solve a large variety of tasks, such as protein function prediction, protein fold classification, structural similarity prediction, and protein-ligand binding affinity prediction. Moreover, we show how fine-tuned networks, pre-trained with our algorithm, lead to significantly improved task performance, achieving new state-of-the-art results in many tasks.