BIOINFORMATICS 101: THE BASICS

After each search round, high-scoring sequences are used to create a multiple alignment that is then used to calculate match scores for the PSSM. When building the PSSMs part of each score is based upon observed amino acid frequencies in the multiple alignment, and part is base on prior knowledge of amino acid substitutability. The prior information, represented as "pseudocounts", is dervied from a standard scoring matrix, such as BLOSUM62. Pseudocounts are particularly useful when the sequences included in the multiple alignments do not constitute an adequate sample of the protein family that they represent. You can control the relative contribution of the alignments and pseudocounts with the pseudocount constant.