POPULATION SEQUENCING FROM CHROMATOGRAM DATA

, gene) from a sample. Traditionally, this is achieved chemically, for example through the use of specific primer sequences. However, it is possible that multiple related species are picked up with the same primer. This is especially problematic in sequencing RNA or proviral DNA, when the virus in question is highly variable and each individual is infected with a different swarm of viral strains. In the case of HIV, when dominant sequences in the population differ by one or more insertions/deletions, standard sequencing techniques fail to recover any component strains satisfactorily. For example, regions of HIV such as the envelope proteins (which are subjected to strong selective pressures from host defenses and result in multiple mutations, insertions, and deletions), most chromatograms obtained by DNA sequencers have unusable sections. We show that chromatograms of mixed sequences, such as the segment shown in Figure 1, can be used to accurately infer the individual strains thus eliminating the need for additional sequencing steps (