Somatic rasiRNAs have not been reported in B. mori, but piRNAs isolated from ovary-derived cell lines show unimodal length distribution with a length peak at 27 nt and a strong bias for U at the 59 end. As reported here, the S. frugiperda LNCR rasiRNAs do not have a prominent nucleotide motif. This is in contrast to some previously described rasiRNAs that interact with the Piwi protein. The unusual features of S. frugiperda LNCR rasiRNAs suggest a somatic origin and Piwiprotein independent biogenesis for these rasiRNAs. Somatic rasiRNAs produced in a PIWI/AGO3 independent manner have already been demonstrated in D. melanogaster COM locus and they operate as silencers of retrotransposons. Because of their size and single-stranded nature, S. frugiperda LNCR rasiRNAs also differ from somatic endo siRNAs whose biogenesis depends on AGO2. Uni-strand rasiRNAs have been reported to arise mainly from the antisense strand of retrotransposons. Therefore, the described LNCR sequence could be the antisense strand of an atypical, probably non-autonomous retrotransposon. S. frugiperda LNCR rasiRNAs are apparently generated from one strand of LNCR and their biogenesis cannot be assigned to ����ping pong���� pairing. We presume that some unknown endonuclease activity is involved in the biogenesis of LNCR rasiRNAs. Depending on the developmental stage, different lengths of S. frugiperda LNCR rasiRNAs were detected. We therefore suppose that the endonucleases involved in their biogenesis could be different and specific to certain developmental stages. At the same time, different secondary structures of two LNCR rasiRNA clusters could indicate different functions. The LNCR rasiRNAs described in this study resemble those described at Drosophila��s pericentromeric flamenco/COM locus. This locus encodes uni-strand rasiRNAs, corresponding to retrotransposon copies dispersed throughout the Drosophila genome and serving as guides, leading to cleavage of expressed retrotransposons in germinal and somatic cells. Families of retroelements distinct from both LTR retrotransposons and non-LTR retrotransposons have been described in many species, and their AG-013736 mechanism of integration is mysterious. The presence of multiple TE-LNCR copies in the S. frugiperda genome means that it has, or used to have the ability to transpose. The absence of homology to transposases or FG-4592 reverse transcriptases in the TE LNCR element and the presence of LNCR transcripts, as putative transposition intermediates, suggest that it belongs to class I of TE and that LNCR may use an unknown reverse transcriptase derived in trans from another genomic source. LNCR RNA could be a retrosequence; a mobile element generated by reverse transcription of mRNA transcripts and subsequent incorporation of the resulting complementary DNA into chromosomal DNA.