PyDescriptorC* and PyDescriptor

In QSAR and machine learning field, molecular descriptor calculators often disappoint when it comes to deal with ease of interpretation and chirality-based descriptors. To solve this, we have developed PyDescriptorC* (earlier known as PyDescriptor). It computes 112,194 molecular descriptors, including 15,150 chirality-based descriptors (~13.5%) for a molecule. All descriptors are easy to interpret and understandable in terms of structural features. It is available from following link: https://sites.google.com/view/pydescriptorcstar/home

Comparative/Benchmark testing with other descriptor calculators:

1. T.B. Kimber, S. Engelke, I.V. Tetko, E. Bruno, G. Godin, Synergy effect between convolutional neural networks and the multiplicity of SMILES for improvement of molecular prediction, arXiv preprint arXiv:1812.04439, (2018).
2. S. Sosnin, D. Karlov, I.V. Tetko, M.V. Fedorov, Comparative study of multitask toxicity modeling on a broad chemical space, Journal of chemical information and modeling, 59 (2018) 1062-1072.