Xiansong Huang

π-SPECFormer: A Mass Spectrometry-based Foundation Model for Decoding the Dark Proteome

This paper introduces π-SPECFormer, a foundation model for mass spectrometry-based proteomics that decodes the dark proteome by pre-training on unannotated spectra and outperforms baselines in peptide sequencing and modification detection.

Predicting protein stability changes upon mutations with dual-view ensemble learning from single sequence

DVE-stability predicts mutation-induced protein stability changes using a dual-view ensemble learning framework, achieving state-of-the-art performance on multiple benchmarks and showing strong generalization to rare beneficial mutations.

Generative prediction of real-world prevalent SARS-CoV-2 mutation with in silico virus evolution

This paper develops a deep-learning framework to predict real-world prevalent SARS-CoV-2 mutations using in silico virus evolution and protein language models, achieving accurate predictions of future dominant mutations.

Interpretable antibody-antigen interaction prediction by introducing route and priors guidance

This paper introduces S3AI, an interpretable method for predicting antibody-antigen interactions by combining structural knowledge and chemical rules, outperforming existing models and identifying key CDR regions.

Peptide generative design with weakly order-dependent autoregressive language model and lifelong learning

This paper introduces PepGenWOA, a generative model for bioactive peptide design that outperforms existing methods and enables efficient virtual screening.

Multiple sequence-alignment-based RNA language model and its application to structural inference

This paper introduces an RNA language model based on multiple sequence alignment, showing it can accurately predict RNA structure from sequence data.

Running ahead of evolution - AI based simulation for predicting future high-risk SARS-CoV-2 variants

This paper develops an AI-based simulation method to predict future high-risk SARS-CoV-2 variants by simulating RBD mutations and screening potential variants efficiently.

Introducing π-HelixNovo for practical large-scale de novo peptide sequencing

The paper introduces π-HelixNovo, a Transformer-based model for de novo peptide sequencing that outperforms existing methods and improves taxonomic resolution in gut metaproteomes.