ProglueDB

Read full text Summarise with VIRAT
Authors: Zhongqi Qi, Nanqi Shen, Yiyan Liao
Cited by: 0

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

ProGlueDB Zenodo Description Draft Title: ProGlueDB: A Curated Dataset for Large-Scale Molecular Glue Virtual Screening Description: Overview ProGlueDB is a comprehensive, high-quality dataset developed as the foundational database for GlueBind, a deep learning framework designed for large-scale molecular glue virtual screening. This dataset was explicitly constructed to address the scarcity of experimentally validated ternary complex data, which has historically hampered the rational design of targeted protein degradation (TPD) agents such as molecular glue degraders (MGDs) and PROTACs. Dataset Contents and Sources The dataset comprises 49,072 high-quality, non-redundant degrader-target entries. To build this comprehensive collection, raw records encompassing diverse degrader-target interaction data were initially aggregated from four publicly available databases: · MGTbind · MolGlueDB · TPDdb · MGDB · Data Curation and Processing · To ensure the reliability of the data for downstream machine learning applications, the raw entries were subjected to rigorous quality control and standardization steps: · Filtering: Entries with inconsistent formatting, missing protein interaction annotations, or those not corresponding to tripartite/ternary complexes were removed. · Decomposition: Entries associated with multiple target proteins were decomposed into individual records to maintain a standardized one-to-one correspondence between a specific degrader and its target. · Supplementation: Incomplete records were supplemented by querying the UniProt database to retrieve canonical sequence and functional information. · Deduplication: Duplicate entries were identified and removed based on exact matches in both molecular structure (canonical SMILES) and target protein identity. · Related Publication · This dataset is officially associated with the following research article: · Title: GlueBind: A Deep Learning Framework for Large-Scale Molecular Glue Virtual Screening · Authors: Yiyan Liao, Nanqi Shen, and Zhongqi Qi

Fields of Study & Subjects

Protein Degradation and Inhibitors Computational Drug Discovery Methods Chemical Synthesis and Analysis

Publication details

Journal / source
Zenodo (CERN European Organization for Nuclear Research)
Publication date
28 March 2026
Language
EN
Document type
Dataset