Category Archives: Opentox

This article illustrates feature generation, feature selection, and validation using opentox-ruby and the REST paradigm. As such, it may be useful as a high-level documentation.  REST (REpresentational State Transfer) is the programming paradigm that HTTP, the WWWs ‘engine’, is based upon. … Continue reading →

Opentox-ruby has selection routines to filter out useful features for predictive models. Features are filtered using non-parametric resampling (bootstrapping) with 500 trees in a random forest. In a preliminary step, the usefulness of every feature is assessed, then subsets of … Continue reading →

Opentox-ruby has facilities to calculate more than 300 non-proprietary physico-chemical descriptors, whose calculation can be easily triggered with REST calls. Three libraries are employed for descriptor calculation: Chemistry Development Kit, developed by Chris Steinbeck, Egon Willighagen, and others. OpenBabel, developed by … Continue reading →

Several statistical learners exist within Lazar. This page details important computational components. Model Building Data Mining Modules Lazar Component Subgraphs PC Descriptors Feature Generation URI* BBRC, LAST‑PM Ambit (several) Feature Generation Algorithm Substructure.match, Substructure.match_hits Substructure.lookup Feature Type nominal or frequency … Continue reading →

Important use-cases and command line arguments / web service switches are presented below. The list is not complete! See fminer README and web services README for a more complete overview. BBRC Use-Case Command Line Web Service API n-Frequent Subgraph Mining (w/o adding … Continue reading →