Published August 30, 2021 | Version 1.0.0
Dataset Open

LibINVENT: Reaction-based Generative Scaffold Decoration for in Silico Library Design

  • 1. Molecular AI, Discovery Sciences, R&D, AstraZeneca, Gothenburg 43183, Sweden
  • 2. Molecular AI, Discovery Sciences, R&D, AstraZeneca, Gothenburg 43183, Sweden Department of Pharmaceutical Biosciences, Uppsala University, Uppsala 75237, Sweden

Description

Training datasets used for LibINVENT publication https://doi.org/10.1021/acs.jcim.1c00469.

The datasets used for training of the prior:

  • purged_chembl_sliced.smi.gz: The CHEMBl 27 compounds, filtered according to the rules described in the manuscript and sliced according to the reaction rules.
  • chembl_train.smi.gz: The purged, sliced dataset used for model training. The DRD2 compounds are removed as described in the manuscript.

Files

Files (1.4 GB)

Name Size Download all
md5:eb5327276fc44dce18c6917285b64b14
693.3 MB Download
md5:eb5327276fc44dce18c6917285b64b14
693.3 MB Download