IDEPI: rapid prediction of HIV-1 antibody epitopes and other phenotypic features from sequence data using a flexible machine learning platform.

摘要: Since its identification in 1983, HIV-1 has been the focus of a research effort unprecedented scope and difficulty, whose ultimate goals — cure vaccine – remain elusive. One fundamental challenges accomplishing these is tremendous genetic variability virus, with some genes differing at as many 40% nucleotide positions among circulating strains. Because this, bases viral phenotypes, most notably susceptibility to neutralization by particular antibody, are difficult identify computationally. Drawing upon open-source general-purpose machine learning algorithms libraries, we have developed software package IDEPI (IDentify EPItopes) for genotype-to-phenotype predictive models from sequences known phenotypes. can apply learned classify unknown also specific sequence features which contribute phenotype. We demonstrate that achieves performance similar or better than previously published approaches on four well-studied problems: finding epitopes broadly neutralizing antibodies (bNab), determining coreceptor tropism identifying compartment-specific signatures deducing drug-resistance associated mutations. The cross-platform Python source code (released under GPL 3.0 license), documentation, issue tracking, pre-configured virtual be found https://github.com/veg/idepi.