Fingerprint-based Compound Classifier

Overview
Ligand classification is an essential component in modern computer-aided drug design. Fingerprint-based compound classifier (FCC) is a novel ligand classification program that implements LiCABEDS (Ligand Classifier of Adaptively Boosting Ensemble Decision Stumps) in Visual C++ .NET 2005. The performance of LiCABEDS is thoroughly discussed in the original manuscript. In FCC, a variety of friendly graphical user interface is provided to automate project management, dataset importing, model training, prediction making, and so on. Our goal is to help pharmacologists and medicinal chemists to easily use machine learning techniques for virtual screening and quantitative structure-activity relationship (QSAR) study.

Description
This is the first release of the FCC program for Windows platform. Executable binary files and sample datasets can be found from the "Download" link. The program is compiled as "green software", so the installation is not required. Please refer to the "readme" file for the usage of each file. Software manual briefly introduces the interface and functions of the FCC program, which also includes a "walk-through" project to give users a general idea of the software.