Genomic evidence for the adaptation of the influenza virus to its human host

The Type A influenza virus (Orthomyxoviridae) is a spherical or rod-shaped virus with a single-stranded, negative-sense RNA genome divided into eight segments that encode eight proteins and two glycoproteins. Thorough analysis of influenza's intergenomic relationship to its human host is vital to improving its therapeutic management. Due to the nature of influenza's genome, such analysis requires an alignment-independent computational protocol focused on short sub-sequence inter-genomic similarities with an efficient capacity for exhaustive mutational analysis. Recently, the University of Houston Bioinformatics Laboratory developed novel computational algorithms to conduct this analysis. The resultant metric of genomic distance (Form-free Distance or F-Distance) is essentially a per-nucleotide average of the number of edits required to render one genome identical to another. F-Distance analysis revealed inter-genomic relationships between the influenza virus and its human host while thoroughly accommodating mutations without detail-limiting heuristic compromises. F-Distances of Type A influenza are not fully attributable to random sequence effect or GC-content bias, and they are not due to the ubiquitous presence of retroelements in the human genome, as demonstrated by host-cohort and serotype-cohort analyses. Further, F-Distance analysis revealed multiple significant, segment-specific temporal trends within the histories of H1N1 and H3N2, and these temporal trends imply the presence of mimicry 'targets' within the human genome. Accordingly, I reversed the analytical foreground and background, comparing segment cohorts of the entire concatenated influenza genome to individual sequences of the human transcriptome. I identified multiple human transcripts with normalized F-Distance means exceeding the 99.99% significance level of the overall normalized human transcriptome mean. Many human transcripts thus identified had known roles in the influenza infection cycle, thereby confirming influenza's mimicry of these human genomic sequences. Finally, during April of 2009, a virulent H1N1 strain of swine-specific Type A influenza virus appeared in Mexico. This unfortunate circumstance presented an opportunity to therapeutically apply F-Distance analysis to an epidemiological crisis for discovery of ultraspecific PCR primers and microarray probes diagnostic for the emerging strain. This analysis identified three suitable oligonucleotides within segment six of the 2009-H1N1 genome, which collaborators clinically validated using RT-PCR. These results demonstrated the capacity of F-Distance analysis to identify the host-dependent nature of influenza evolution, which both mimics and avoids the host genome through time via mutational acquisition of distributed short sub-sequence similarities.