Crossmap liftover: Chromosome changed between hg38 and hg19???

Question

[skim]

I have done a liftover on a hg38 bed file to a hg19 bed file on Crossmap(Link) and found out that multiple queries were changed from chromosome 22 to chromosome 14. Should I just filter this or did I do something wrong?

Thank you very much.

Data: hg38

chr22   15690509    15690536    POTEH   28.8238274624   -
chr22   15698714    15698741    POTEH   69.8367597081   -
chr22   15698718    15698745    POTEH   63.510218446    +
chr22   15698719    15698746    POTEH   64.7710373827   +
chr22   15698720    15698747    POTEH   71.7040268333   -
chr22   15698720    15698747    POTEH   64.4982928966   +

hg19

chr14   19553737    19553764    POTEH   28.8238274624   -
chr14   19562046    19562073    POTEH   69.8367597081   -
chr14   19562052    19562079    POTEH   71.7040268333   -

Some of these were already filtered based on 27bp length. All of the data are 27bp. Chain file is this: http://hgdownload.soe.ucsc.edu/goldenPath/hg38/liftOver/hg38ToHg19.over.chain.gz

Answer 1

[Devon Ryan]

You didn't do anything wrong, if you zgrep chr22 hg38ToHg19.over.chain.gz | grep chr14 you'll see there's a bunch of sequence originally in chr14 that was moved to chr22. If you go over to the gencode website and scroll through the GRCh37 assembly issues, you'll see a LOT of them were fixed with the GRCh38 release.