Maybe you are observing the phenomenon described initially by my colleagues from Ceske Budejovice: https://www.researchgate.net/publication/224867908_A_widespread_occurrence_of_extra_open_reading_frames_in_plant_Ty3gypsy_retrotransposons Otherwise, I would suspect an error caused by genome rearrangements after insertion. You should carefully inspect the individual cases and try to come up with what these rearrangements might have been.

Article A widespread occurrence of extra open reading frames in plan...

Thank you Matej.

Almost anything is possible with LTR based transposons. The problem with any sort of automated identification of transposons is not with finding them but working out what it all means.  You may have inadvertently identified two LTRS from completely unlinked transposons but just because they are within the genomic distance from each other that you have defined in your search parameters the software erroneously reports them as belonging to the same locus. the long and the short of it is that you will probably need to do some sort of secondary quality control and screening (and in the end it often boils down to eyeballing the sequences) to make sense out of it all 

Dear Swaraj Basu

My experience with programmes that search LTR retrotransposons in genomes suggest me that probably you are looking two LTR retrotransposons belonging to the same family but proxime between them (and flanking a gene). Probably if you search about 3000 or 5000 bp upstream and downstream each LTR you can find their respective LTR. For me, this is the most probable explanation.

Good luck in your research. Best regards, Rosalia