Principal component based method for whole genome phylogenetic analysis without alignment: Application to HEV genotype

We describe principal component method for the DNA sequence analysis using digital filters. With the huge amount of data accessible in the public domain, digital filters are very helpful in DNA sequence processing. In this technique, the occurrence frequency of the q-gram genetic word of interest is determined from the DNA sequence. The sequence is then elucidated by using finite impulse response (FIR) type filter in order to determine the q-gram word density along the sequence. The word density distribution is further used for principal component analysis (PCA) to determine the similarity / dissimilarity between the sequences. The technique is verified by using 48 HEV genotypes. The results are in good agreement with other methodology.