HiTE: An accurate dynamic boundary adjustment approach for full-length Transposable Elements detection and annotation in Genome Assemblies

Abstract

Recent advancements in genome assembly have greatly improved the prospects for comprehensive annotation of Transposable Elements (TEs). However, existing methods for TE annotation using genome assemblies are less accurate and robust, requiring extensive manual editing. In addition, the currently available gold-standard TE databases are not comprehensive, even for extensively studied species, highlighting the critical need for an automated TE detection method to supplement existing repositories. In this study, we introduce HiTE, an accurate dynamic boundary adjustment approach designed to detect full-length TEs. The experimental results demonstrate that HiTE identified 932 perfect TE models with a precision of 0.971 on the rice reference genome, which are 142% and 4.42% higher than the state-of-the-art tool RepeatModeler2, respectively. Furthermore, HiTE discovers over 800 novel TIR elements with well-defined structures that are not included in known libraries, enabling the discovery of new insights. We have also implemented a Nextflow version of HiTE to enhance its parallelism reproducibility, and portability.