Covariance Search Model for Identifying ncRNA using Particle Swarm Optimized Agglomerative Clustering

Abstract

Recent studies have shown that the functional discovery of noncoding ribonucleic acids (ncRNA) is gradually gaining interest among bioinformatics experts. Families of ncRNAs are responsible for various biological functions, including gene expression regulation and catalytic activities, which have yet to be discovered. These discoveries have expanded the scope of the ncRNA study, including finding functional subgroups. Hence, cross-fertilisation solutions derived from computational intelligence principles and algorithms have begun to produce promising outcomes. For the Covariance Model (CM) in ncRNA identification, data clustering is one of the most common strategies in various fields. Based on sequence similarity, hierarchical clustering is the most common method for classifying a set of human ncRNAs into distinct families. However, standard techniques have several drawbacks, such as the sequence structures of each family getting considerably diluted as the number of sequence characteristics in the known family dataset grows. This study optimises the hierarchical clustering approach for identifying ncRNA families using Particle Swarm Optimization (PSO).