Enregistré dans:
Détails bibliographiques
Auteurs principaux: Chithari, Vivek Reddy, Young, Jasmine Y., Persikova, Irina, Liang, Yuhe, Crichlow, Gregg V., Flatt, Justin W., Ghosh, Sutapa, Hudson, Brian P., Peisach, Ezra, Sekharan, Monica, Shao, Chenghua, Burley, Stephen K.
Format: Preprint
Publié: 2026
Sujets:
Accès en ligne:https://arxiv.org/abs/2604.22800
Tags: Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
_version_ 1866908991371608064
author Chithari, Vivek Reddy
Young, Jasmine Y.
Persikova, Irina
Liang, Yuhe
Crichlow, Gregg V.
Flatt, Justin W.
Ghosh, Sutapa
Hudson, Brian P.
Peisach, Ezra
Sekharan, Monica
Shao, Chenghua
Burley, Stephen K.
author_facet Chithari, Vivek Reddy
Young, Jasmine Y.
Persikova, Irina
Liang, Yuhe
Crichlow, Gregg V.
Flatt, Justin W.
Ghosh, Sutapa
Hudson, Brian P.
Peisach, Ezra
Sekharan, Monica
Shao, Chenghua
Burley, Stephen K.
contents Motivation: Structural Biologists have contributed more than 245,000 experimentally determined three-dimensional structures of biological macromolecules to the Protein Data Bank (PDB). Incoming data are validated and biocurated by ~20 expert biocurators across the wwPDB. RCSB PDB biocurators who process more than 40% of global depositions face increasing challenges in maintaining efficient Help Desk operations, with approximately 19,000 messages in approximately 8,000 entries received from depositors in 2025. Results: We developed an AI-powered Help Desk using Retrieval-Augmented Generation (RAG) built on LangChain with a pgvector store (PostgreSQL) and GPT-4.1-mini. The system employs pymupdf4llm for Markdown-preserving PDF extraction, two-stage document chunking, Maximal Marginal Relevance retrieval, a topical guardrail that filters off-topic queries, and a specialized system prompt that prevents exposure of internal terminology. A dual-LLM architecture uses separate model configurations for question condensing and response generation. Deployed in production on Kubernetes with PostgreSQL (pgvector), it provides around-the-clock depositor assistance with citation-backed, streaming responses. Availability and implementation: Freely available at https://rcsb-deposit-help.rcsb.org.
format Preprint
id arxiv_https___arxiv_org_abs_2604_22800
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle RCSB PDB AI Help Desk: retrieval-augmented generation for protein structure deposition support
Chithari, Vivek Reddy
Young, Jasmine Y.
Persikova, Irina
Liang, Yuhe
Crichlow, Gregg V.
Flatt, Justin W.
Ghosh, Sutapa
Hudson, Brian P.
Peisach, Ezra
Sekharan, Monica
Shao, Chenghua
Burley, Stephen K.
Information Retrieval
Artificial Intelligence
Computation and Language
Quantitative Methods
Motivation: Structural Biologists have contributed more than 245,000 experimentally determined three-dimensional structures of biological macromolecules to the Protein Data Bank (PDB). Incoming data are validated and biocurated by ~20 expert biocurators across the wwPDB. RCSB PDB biocurators who process more than 40% of global depositions face increasing challenges in maintaining efficient Help Desk operations, with approximately 19,000 messages in approximately 8,000 entries received from depositors in 2025. Results: We developed an AI-powered Help Desk using Retrieval-Augmented Generation (RAG) built on LangChain with a pgvector store (PostgreSQL) and GPT-4.1-mini. The system employs pymupdf4llm for Markdown-preserving PDF extraction, two-stage document chunking, Maximal Marginal Relevance retrieval, a topical guardrail that filters off-topic queries, and a specialized system prompt that prevents exposure of internal terminology. A dual-LLM architecture uses separate model configurations for question condensing and response generation. Deployed in production on Kubernetes with PostgreSQL (pgvector), it provides around-the-clock depositor assistance with citation-backed, streaming responses. Availability and implementation: Freely available at https://rcsb-deposit-help.rcsb.org.
title RCSB PDB AI Help Desk: retrieval-augmented generation for protein structure deposition support
topic Information Retrieval
Artificial Intelligence
Computation and Language
Quantitative Methods
url https://arxiv.org/abs/2604.22800