Abstract: LTR retrotransposons are an important class of eukaryotic transposable elements, which are ubiquitous and highly heterogeneous in plant and play a major role in genome evolution of eukaryote. They are now extensively employed in gene function and genetic diversity analyses. Identification of LTR retrotransposons is the precondition for its application. Therefore, it has important theoretical significance and practical application value in studying identification and analysis methods LTR retrotransposon sequences. Bioinformatic software of the sequence analysis, according to the work principle, can be classified roughly into two types: sequence alignment and sequence identification of conserved domains. Alignment software, such as BLAST and DNAstar, produce the corresponding sequence information through comparison of sequence similarity; however, this kind of software cannot be applied for full length sequences. According to the principle, LTR retro-transposon sequence identification software can be roughly sorted into four types: de novo repeat discovery method, comparative genomic method, homology-based method, and structure-based method. For example, LTR_Finder based on de novo repeat discovery method can accurately predict and annotate LTR retrotransposons for full length sequences; RepeatMasker, which is based on homology-based method, can discover LTR retrotransposons by comparing the similarity with known sequences in the database. In this article, different methods of identification and analysis of retrotransposon sequences were compared and analyzed, and a set of flow of LTR retrotransposons sequence analysis was summarized in order to provide the reference for LTR retrotransposons sequence analysis.

Key words: plant LTR retrotransposons, sequence alignment, software analysis, genetic diversity