eliochen / ochem-data

published organic chemistry data. This is a fork of original repo (link below). Automatically pulled by bot.

ochem-data

This repo compiles published organic chemistry data, including raw data and calculated descriptor sets for molecules and reactions.

File structure:

Each folder contains one reaction (see publication list for details).

Every reaction folder was divided into two sub-folders:

• mols: a list of molecules categorized by reaction roles)
• rxns: a list of reaction entries with multiple reaction components and yield).

Reading data directly from this repo

import pandas as pd

REPO_PATH = 'https://raw.githubusercontent.com/beef-broccoli/ochem-data/main/'
FP = 'deoxyF/paper-dft/train.csv'  # change this  
df = pd.read_csv(REPO_PATH + FP)

# do things with df...

Each subfolder (ohe, mol2vec, mordred...) includes different descriptor encodings for the list of molecules or reaction entries

Publications for reaction data:

• deoxyF: Deoxyfluorination with Sulfonyl Fluorides: Navigating Reaction Space with Machine Learning

• Vbur: Linking Mechanistic Analysis of Catalytic Reactivity Cliffs to Ligand Classification

• CN: Predicting reaction performance in C–N cross-coupling using machine learning

• asym_epox: Ni/Photoredox-Catalyzed Enantioselective Cross-Electrophile Coupling of Styrene Oxides with Aryl Iodides

• CH-aryl-conditions: Bayesian reaction optimization as a tool for chemical synthesis

• kraken: A Comprehensive Discovery Platform for Organophosphorus Ligands for Catalysis

• NiB: Advancing Base Metal Catalysis through Data Science: Insight and Predictive Models for Ni-Catalyzed Borylation through Supervised Machine Learning

Publications/resources for descriptor sets:

• Mordred: a molecular descriptor calculator
• Mol2vec: Unsupervised Machine Learning Approach with Chemical Intuition
• alvaDesc
• Kraken: A Comprehensive Discovery Platform for Organophosphorus Ligands for Catalysis
• auto-qchem