overlapping genes in lncrna prediction

Question

[dineshtripathy9658]

i am following a pipeline for rice lncrna identification. first i removed those with basepair length below a particular point. . then those id above a particular orf bp. then found matches to swissprot and removed the rest. then then removed those have protein coding potential .for the lncrna identification. .i did a windowbed for window 500 from genomic region both in upstream and downstream. removed those id which were overlapping. even then i got LOC_Os*g* id when i fetched data for the final lncrna count from gene_exp.diff. can anyone explain how??