Viral Quasispecies Reconstruction Based on Unassembled Frequency Estimation

The genomic diversity of RNA viruses (such as Hepatitis C virus (HCV), Human immunodeficiency virus (HIV), SARS and influenza) is a subject of the great interest since it is a plausible cause of vaccines failures and virus resistance to existing therapies. RNA lacks ability to detect and repair mistakes during replication, many mutations are well tolerated and passed down to descendants producing a family of co-existing related variants of the original viral genome referred to as quasispecies [4, 14, 11]. Knowing the sequences of the most virulent variants can help in the design of effective drugs [3, 13] and vaccines [7, 5] by targeting particular viral genome in vivo. This paper is devoted to the following problem. Quasispecies Spectrum Reconstruction (QSR) Problem. Given a collection of 454 pyrosequencing reads taken from a sample quasispecies population, reconstruct the quasispecies spectrum, i.e., the set of sequences and the relative frequency of each sequence in the sample population.