Welcome to NanoBind!

         Nanobodies not only have the ability to specifically bind antigens, but also exhibit unique small-molecule characteristics, low immunogenicity, and strong tissue penetration. As the subset of protein-protein interactions (PPIs), nanobody-antigen interactions (NAIs) hold significant importance in elucidating immune mechanisms and the de novo design of nanobodies. NanoBind is a sequence-based deep learning model designed to predict NAIs, antigen-binding sites, affinity values, and affinity strength comparisons.

         NanoBind directly utilizes amino acid sequences of proteins as input, employs the pre-trained language model ESM-2 for embedding, and extracts features through a self-attention mechanism with rotary position embeddings and convolutional neural networks to generate sequence representations.

         We provide researchers with a convenient and efficient NanoBind webserver for predicting a range of nanobody-antigen interaction tasks. Additionally, you can access the source code of NanoBind on our GitHub page https://www.bilibili.com/.

If you think NanoBind is useful, please kindly cite the following paper:

Webserver update: