Prediction of the next mutation in Hemagglutinin protein of Influenza-A virus using the variation pattern analysis.

Every year certain treatment strategies are developed to combat influenza that too only after the epidemic hits the population. The major setback in designing the treatment strategy is due to the variation in the surface antigenic determinants (Hemagglutinin and Neuraminidase) of the virus. In this work, the position specific contribution of an amino acid in the variation of the Hemagglutinin protein has been derived. Multiple sequence alignment of non-redundant sequences of Hemagglutinin from different strains has been used to derive a position specific weighted probability score matrix. The next-in-line variation in the subtype of the Hemagglutinin protein of the influenza A virus has been predicted using the calculated score matrix. Although the prediction has been accomplished with an average accuracy of 60%, the accuracy can still be improved. This strategy may be proven to be useful to design a drug before the outburst of the disease.