ConvPred: A deep learning-based framework for predictions of potential organic reactions

dataset

posted on 2024-06-28, 02:30 authored by Wenlong WangWenlong Wang

The "Codes.zip" contains python scripts for training sigma-profiles-based and ECFP4-based models and script for generating ECFP4 dict.

The "Models.zip" contains trained sigma-profiles-based and ECFP4-based models using the scripts mentioned above.

The "Data.zip" constains reactions used for training and their sigma-profiles descriptor generated by MOPAC.

The "VRF(sigma).7z" contains the python script used to generate certain VRF(sigma) for chemical reactions.

The "Scripts for extracting reactions and templates.7z" contains the python scripts used to extract certain types of reactions and templates from the USPTO 1976-2016 database.

Funding

National Natural Science Foundation of China (22078041)

National Natural Science Foundation of China (22278053)

Fundamental Research Funds for the Central Universities (DUT22LAB608)

Dalian High-level Talents Innovation Support Program (2021RQ105)

History

Usage metrics

Keywords

Reaction Prediction ConvPred Deep Learning USPTO_DUT Cheminformatics Chemical Engineering Design Knowledge Representation and Machine Learning

Licence

CC BY 4.0