Data, Models, and Python Code For: Machine learning models and performance dependency on 2D chemical descriptor space for retention time prediction of pharmaceuticals

Published: 31 May 2024| Version 1 | DOI: 10.17632/x925rnxzcb.1
Contributor:
Armen Beck

Description

Training data, models, and python code for the manuscript: Machine learning models and performance dependency on 2D chemical descriptor space for retention time prediction of pharmaceuticals. MOE descriptors of the METLIN SMRT dataset (original by Domingo-Almenara et. al. with RTs and structures available at: https://figshare.com/ndownloader/files/18130628), training scripts (python), UMAP, GMM, and SVR models with training splits results are within the SMRT_exp.zip file. CSVs for feature importance for SVR models are standalone files.

Files

Institutions

Merck

Categories

Machine Learning, Cheminformatics

Licence