Prediction and annotation of genomic repeat dynamics in the snail <i>Biomphalaria glabrata</i> using Hidden Markov Models

Abstract

Mobile elements cover an extensive amount of the genomes of both plants and animals. However, current homology, or similarity comparison, based search tools are optimized only for analyzing and annotating repeats in humans and well known experimental models. This skewed taxonomic distribution of reference data makes homology-based search tools less sensitive and less accurate, missing many targets in poorly examined genomically diverse lineages. With these limitations in mind, Hidden Markov Models (HMMs) were used for <i>de novo</i>, rather than homology-based, repeat annotation in the gastropod mollusk species <i>Biomphalaria glabrata</i>. Here we compare the HMM profile repeat annotation output to the commonly used homology-based search tool RepeatMasker in order to assess comparatively whether there is an advantage of using <i>de novo</i> model-driven repeat annotation methods over homology-based tools. Finally, we used PCR to amplify eight choice repeat segments across five cloned individual <i>B. glabrata</i> to verify experimentally the output received computationally.