An algorithm for mapping positively selected members of quasispecies-type viruses

Area of Science:

• Virology
• Bioinformatics
• Computational Biology

Background:

• RNA viruses exist as quasispecies, a collection of related variants.
• Viral sequence variability poses significant bioinformatics challenges.
• Analyzing the complete mutational spectrum is crucial for understanding viral genomes.

Purpose of the Study:

• To develop a novel algorithm for identifying positively selected variants in viral proteins.
• To address the challenge of analyzing sequence variability in viral quasispecies.
• To provide a biologically relevant representation of viral mutational spectra.

Main Methods:

• Development of the "selection mapping" algorithm, named QUASI.
• Identification of replacement mutations overabundant relative to silent mutations at each codon.
• Distinguishing positively selected mutations from neutral or negatively selected "noise".

Main Results:

• QUASI successfully identifies positively selected variants of viral proteins.
• The algorithm pinpoints specific amino acid changes conferring selective advantage.
• Selection mapping effectively differentiates beneficial mutations from background noise.

Conclusions:

• Selection mapping represents a fundamental improvement over existing methods like dN/dS.
• QUASI's selection maps provide detailed characterization of selected mutational landscapes.
• The study characterized the selected mutational landscapes of influenza A H3 hemagglutinin, HIV-1 reverse transcriptase, and HIV-1 gp120.