Clustering using Positional Association Rules Algorithm on Protein Sequence Motifs

In the field of Bioinformatics, protein sequence motif analysis has the potential to discover the role and function of specific proteins. Many algorithms or techniques to discover motifs require a predefined fixed window size in advance. Due to the fixed size, these approaches often deliver a number of similar motifs simply shifted by some bases or including mismatches. For this problem, earlier work in this field has produced a new Association Rule aptly named the Positional Association Rule, which searches for frequent distances appearing among motifs and motifs. In this work we adapt this newly proposed association rules algorithm to “cluster” the protein sequence motifs to solve the problem generated by fixed window size approach. Although association rule based algorithms have been widely adapted in association analysis and classification, few of those are designed as clustering methods. We believe this paper will provide future research a novel way to find patterns and associations when dealing with protein sequences or similar applications.