Extremely long RefSeq sequence for miRNA

Question

[knowah]

I was taking a look at which RefSeq genes (hg19) intersected a list of 200bp genomic regions, and found a peculiar entry: MIR548H4 (RefSeq NR_031680) is located at chr15:69116303-69489862, which is about 300kb long. Other (non-RefSeq) sequence coordinates for this miRNA are on chr8 and about 100bp long (NCBI, GeneCards on UCSC).

I checked the RefSeq refGene list from UCSC and found 83 miRNA's which had a sequence of over 200bp. Why are there micro RNA's in RefSeq that are so long (especially when "the RefSeq represents the predicted [~70bp] microRNA stem-loop")? And is the MIR548H4 located on the wrong chromosome by accident?

Answer 1

[Devon Ryan]

That's a funky refseq entry. If you look at it in depth, you'll see ~24bp aligns at one end and ~32 at the other. It looks like the entry is simply wrong.

Edit: You'll see this is fixed in the hg38 annotation.