A Computer Program Designed to Screen Rapidly for HIV Type 1 Intersubtype Recombinant Sequences

D etection of recombinant genomes involving the genetically distinct HIV-1 subtypes is a current problem of immense importance1-3; if such intersubtype recombination occurs with notable frequency, the already formidable challenge of devel- oping vaccines against multiple subtypes is further complicated.Here we describe a simple and rapid heuristic method for de- tection of recombinant sequences (or more broadly, mosaic se- quences), and its application to the env and gag coding se- quences in the HIV Sequence Database.4The computer program developed, called the Recombinant Identification Program (RIP), quickly provides summary infor- mation describing large sets of sequences (e.g., Table 1) and more detailed output describing particular sequences (e.g., Fig. 1).Two sequence sets with compatible alignments are required as input: a "background" alignment with subtype designations (currently HIV-1 subtypes A-H have been defined4), and an alignment of query sequences to be screened one at a time.Alignments may be either of nucleotide or amino acid se- quences.The RIP first generates a consensus sequence to rep- resent each subtype as delineated in the background alignment.A consensus threshold option allows the user to select a mini- mal frequency for the most common character in a given posi-