100% Python-based QSAR package that utilizes random forest, neural network, consensus voting, autoML, and LMST to predict property based on structure (quntitative structure analysis relationship).
- Random forest
- Neural network
- Combinatorial QSAR (consensus voting)
- Williams Plot applicability domain
- autoML framework using TPOT
- Saving/loading descriptors (generated from Mordred)
- Descriptor-free regression model using LMST
- Data wrangling and preprocessing