The invention relates to the technical field of
bioinformatics, in particular to a method for identifying
RNA (Ribonucleic Acid) from a
transposon element based on
RNA sequencing data. According to the invention, the
exon level of the teRNA is taken as an analysis object, compared with a teRNA site, the teRNA
exon not only realizes more accurate site identification, but also retains the existing form of TE in a transcription product, including whether a splicing site and a
polyadenylic acid signal are provided or not, whether the TE is chimeric with a
gene or other TE sites or not and the like. And compared with the level of the teRNA transcript, as a constituent part of the transcript, the teRNA
exon significantly reduces the
sequence assembly difficulty of the short-read-length
sequencing data. According to the method, the Bayesian model, the parametric
assembly algorithm and the local site de novo
assembly algorithm are combined, and through mutual correction of different
assembly algorithms and repeated comparison sequencing reading length before and after assembly, the assembly false positive is reduced, so that the teRNA recognition accuracy is improved.