Evaluating parameters for ligand-based modeling with random forest on sparse data sets

• Published in: Journal of cheminformatics 2018-10, Vol.10 (1), p.49-10, Article 49
• Main Authors: Kensert, Alexander, Alvarsson, Jonathan, Norinder, Ulf, Spjuth, Ola
• Format: Article
• Language: English
• Subjects: Chemical properties; Chemistry; Chemistry and Materials Science; Computational Biology/Bioinformatics; Computer Applications in Chemistry; Data analysis; Data processing; Data structures; Datasets; Decision making; Decision trees; Documentation and Information in Chemistry; Fingerprint; Fingerprints; Learning algorithms; Ligands; Ligands (Chemistry); Machine learning; Mathematical models; Memory; Modelling; Parameters; Performance prediction; Prediction models; Random forest; Research Article; Software reviews; Sparse representation; Support vector machines; Theoretical and Computational Chemistry
• ISSN: 1758-2946; 1758-2946
• DOI: 10.1186/s13321-018-0304-9

Ligand-based predictive modeling is widely used to generate predictive models aiding decision making in e.g. drug discovery projects. With growing data sets and requirements on low modeling time comes the necessity to analyze data sets efficiently to support rapid and robust modeling. In this study we analyzed four data sets and studied the efficiency of machine learning methods on sparse data structures, utilizing Morgan fingerprints of different radii and hash sizes, and compared with molecular signatures descriptor of different height. We specifically evaluated the effect these parameters had on modeling time, predictive performance, and memory requirements using two implementations of random forest; Scikit-learn as well as FEST. We also compared with a support vector machine implementation. Our results showed that unhashed fingerprints yield significantly better accuracy than hashed fingerprints ( p ≤ 0.05 ), with no pronounced deterioration in modeling time and memory usage. Furthermore, the fast execution and low memory usage of the FEST algorithm suggest that it is a good alternative for large, high dimensional sparse data. Both support vector machines and random forest performed equally well but results indicate that the support vector machine was better at using the extra information from larger values of the Morgan fingerprint’s radius.