Fandi Wu

FoldAbS repurposes protein folding models as foundation encoders to improve antibody screening by evaluating specificity, affinity, and developability in a unified framework.

This paper introduces InterAb and InterAb-Opt, computational frameworks that predict and optimize broadly neutralizing antibodies using all-atom modeling and pre-trained language models, showing significant improvements in binding affinity and neutralization activity.

IgGM2 is an all-atom foundation model for immune receptor design that jointly predicts CDR sequences and full-atom structures, outperforming existing methods in structure prediction and sequence design benchmarks.

This study identifies structural features of affinity-matured R1-32-like antibodies that enable tolerance to SARS-CoV-2 mutations and reveals an ultrapotent antibody with broad variant activity.

The paper uses AI to identify convergent somatic hypermutation patterns in antibodies, enabling the discovery of variant-resilient broadly neutralizing antibodies from early-pandemic data.

This paper introduces TFDesign-sdAb, a deep learning framework for designing therapeutic single-domain antibodies with tailored functionalities, achieving high success rates in generating Protein A-binding sdAbs.

IgGM is a generative model for antibody design that produces high-affinity antibodies through tasks like structure prediction and affinity maturation, validated on multiple antigens.

The paper introduces tFold-TCR, a model for predicting TCR-pMHC structures, achieving higher accuracy and speed than existing methods and creating a large database of these structures.

This paper introduces HuDiff, a diffusion model for humanizing antibodies and nanobodies, showing improved similarity to experimentally humanized sequences and enhanced binding affinity.

IgGM is a generative model that designs functional antibodies and nanobodies by predicting sequences and structures simultaneously for a given antigen.

This paper introduces tFold-Ab and tFold-Ag for fast and accurate prediction of antibody and antibody-antigen complex structures, showing improved speed and performance over AlphaFold-Multimer.

RaptorX-Single predicts protein structures from single sequences using language models and a structure generation module, outperforming MSA-based methods in specific cases.

tFold-Ab predicts antibody structures end-to-end without sequence homologs, achieving accuracy for both backbone and side-chain conformations.

This paper introduces NDThreader, a deep learning method for template-based protein structure prediction that outperforms existing methods on CASP13 and CAMEO data.