The genetically diverse HIV-1 Group M infecting subtypes can be observed as unique branches on a phylogenetic tree and arose due to independent cross-species transmissions between non-human primates and humans. As the HIV-1 pandemic has evolved, different infecting subtypes have prevailed in different geographic populations. The complex factors associated with the global establishment of specific subtypes remains largely unknown. The HIV-1 accessory protein Nef, demonstrates considerable genetic variability and several studies suggest that Nef variation is associated with disease progression. Here we use an evolved neural network approach applied to a well-curated database of HIV-1 Nef sequences from subtypes A1, C, and D, the most prominent subtypes in Uganda, Africa to elucidate functional properties associated with subtype diversity. Following the generation of over 1000 features associated with amino acids physicochemical properties, we use statistical pruning and evolved neural networks to identify key Nef features associated with subtype differentiation. As interest in Nef continues to grow in the research community, we hope that these features foster new understanding of the mechanisms associated with the spread of HIV -1 subtypes in populations.

HIV Nef protein has a known association with cancer, yet the specific features associated with this relationship remain poorly characterized. Using a dataset of well curated Nef sequences, we trained evolved neural networks to classify sequences as having originated from cancer or non-cancer samples. This process allows us to identify features that are possibly associated with Nef protein in HIV and its relation to cancer. We review the methods and features identified for their biological significance.

HIV-1 infects a variety of cell types such as macrophages, T-cells and dendritic cells by expressing different chemokine receptors. R5 HIV-1 viruses use the CCR5 co-receptor for entry, X4 viruses use the CXCR4 co-receptor, and several viral strains make use of both co-receptors (a so-called “dual tropic” or R5X4 virus). Both X4 and R5X4 viruses are associated with late stage rapid progression to AIDS. It remains difficult to identify viral co-receptor type in advance of treatment, especially the R5X4 variety. In this paper we extended previous work to classify HIV-1 tropism using evolved neural networks and a larger set of HIV-1 sequences and features to improve overall classification accuracy.

The HIV-1 genome is highly heterogeneous. This variation affords the virus a wide range of molecular properties, including the ability to infect cell types, such as macrophages and lymphocytes, expressing different chemokine receptors on the cell surface. In particular, R5 HIV-1 viruses use CCR5 as co-receptor for viral entry, X4 viruses use CXCR4, whereas some viral strains, known as R5X4 or D-tropic, have the ability to utilize both co-receptors. X4 and R5X4 viruses are associated with rapid disease progression to AIDS. R5X4 viruses differ in that they have yet to be characterized by the examination of the genetic sequence of HIV-1 alone. In this study, a series of experiments was performed to evaluate different strategies of feature selection and neural network optimization. We demonstrate the use of artificial neural networks trained via evolutionary computation to predict viral co-receptor usage. The results indicate identification of R5X4 viruses with predictive accuracy of 75.5%.

BACKGROUND: Human immunodeficiency virus (HIV) research involves ongoing, repetitious sequencing of the HIV genome and the massive accumulation of associated investigational data. As a result, the storage of annotated DNA and/or protein sequences, as well as information retrieval, have become increasingly difficult tasks, with scientists extracting less information from their collected data than they should. OBJECTIVES: Our objective was to design and develop a software package to aid researchers in the storage, analysis and exploration of their HIV-associated data. RESULTS: HIVbase contains familiar, easy-to-use interfaces and functionality for integrating many types of disparate data. The software contains tools that allow for the mass import of raw genetic data, eliminate repetitious sequence translations, have the ability to identify automatically and store HIV regions of interest from nucleic acid or protein sequences, allow for the export of data in commonly used analysis-ready formats, and for unique querying approaches.