SS-Pro: a simplified Siamese contrastive learning approach for protein surface representation
摘要
In this paper, we introduce a simple Siamese contrastive self-supervised learning framework for protein surface representation learning. The encoder in this framework can be adapted to various point cloud feature extraction backbone networks. Experiments show pre-trained networks consistently demonstrate performance improvements in two downstream tasks. In future work, we aim to explore more efficient protein surface feature extraction networks and delve into additional downstream tasks that better capture protein surface characteristics.