Improving Profile-Profile Alignments via Log Average Scoring

Alignments of frequency profiles against frequency profiles have a wide scope of applications in currently used bioinformatic analysis tools ranging from multiple alignment methods based on the progressive alignment approach to detecting of structural similarities based on remote sequence homology. We present the new log average scoring approach to calculating the score to be used with alignment algorithms like dynamic programming and show that it significantly outperforms the commonly used average scoring and dot product approach on a fold recognition benchmark. The score is also applicable to the problem of aligning two multiple alignments since every multiple alignment induces a frequency profile.