xBind: An LLM-powered webserver for cross-molecular protein binding site prediction
By
Wang, Xinyu, Feng, Xingyue, Tarafder, Sumit, Bhattacharya, Debswapna
Kettled twice. Extra chewy, extra trustworthy.
Summary
xBind is a freely accessible webserver that uses large language model (LLM) embeddings from ESM-2 combined with sequence- and structure-derived features to predict protein binding sites across different molecular types (protein-protein, protein-DNA, and protein-RNA). It employs symmetry-aware deep graph neural networks and accepts input as single-chain protein sequences (FASTA) or monomer protein structures (PDB/mmCIF), outputting predicted residue-level binding site information.
Key quotes
· 3 pulledxBind is an interactive, freely accessible, and fully configurable webserver for large language model (LLM)-enabled cross-molecular protein binding-site prediction.
xBind leverages LLM embeddings from the ESM-2 model together with sequence- and structure-derived features to predict protein–protein, protein–DNA, and protein–RNA binding sites using symmetry-aware deep graph neural networks.
The input to xBind is either a single-chain protein sequence in FASTA format or a monomer protein structure in PDB or mmCIF format and it outputs predicted residue-level bin
You might also wanna read
Analyzing Redundancy in Natural Protein Fold Space and the Role of Generative AI in Biomolecular Modeling
This research note from Ligo explores the concept of redundancy in natural protein fold space, examining how deep neural networks and genera
research.ligo.bio·7d ago
Open-Source LLM Safety Vulnerabilities: How Chat Template Formatting Gates Alignment in Models Like Gemma and Qwen
This article reveals a critical vulnerability in open-source large language models (LLMs) where safety alignment can be bypassed by simply o
teendifferent.substack.com·4mo ago
Using LLMs to Develop a Cross-Platform Elixir NIF in C: Challenges and Insights
The article details the author's experience using large language models (LLMs) like Grok 3, Gemini 2.5 Flash, and GPT-5 to develop a cross-p
Epicure: Multilingual Ingredient Embeddings from 4.14M Recipes Using Skip-Gram and Metapath2Vec
Epicure is a research project that develops three sibling skip-gram ingredient embeddings trained on a multilingual recipe corpus of 4.14M r
arxiv.org·14d ago
Fast-dLLM: Training-Free Acceleration Method for Diffusion Language Models Using KV Cache and Parallel Decoding
Researchers introduce Fast-dLLM, a training-free acceleration method for diffusion-based large language models that addresses their slower i
arxiv.org·7mo ago
Orthrus: A Dual-Architecture Framework for Fast, Lossless LLM Inference via Diffusion Decoding
Orthrus is a dual-architecture framework that combines autoregressive LLMs with diffusion models to enable fast, lossless parallel token gen
github.com·25d ago