sequences should be the same right? only the coordinates will change.
Hi!
Yes, I know about the Lift over tool from UCSC. The problem here is bit more complicated.
I have a raw BED file in hg18, that looks like this:
chr1 235 535 AACCCTAACCCTAAACCCTAAACCCTAATCCC 1 +
chr1 2913 3213 GTCGACCCAGCTGGAGGCAGCCAGGCTTGCCT 1 -
chr1 2921 3221 CTGGCTGCCTCCAGCTGGGTCGACAGACAGGG 1 +
chr1 2925 3225 CTGCCTCCAGCTGGGTCGACAGACAGGGGCTG 1 +
chr1 2934 3234 GCTGGGCCGACAGACAGGGGCTGGAGAAGGGG 1 +
chr1 3160 3460 ACCAGGGCTCAGATCCAGGCCAACAGCGTGTC 1 -
chr1 3162 3462 CACGCTGTTGGCCTGGATCTGAGCCCTGGTGG 1 +
chr1 3167 3467 GACCTCCACCAGGGCTCAGATCCAGGCCAACA 1 -
chr1 3169 3469 TTGACCTCCACCAGGGCTCAGATCCAGGCCAA 1 -
chr1 3173 3473 GGCTTTGACCTCCACCAGGGCTCAGATCCAGG 1 -
I want to convert this file into hg19.
Now, if I had the FASTQ file for this one it would be a regular exercise. But I don't have it . Secondly, liftover will only convert first three columns and the other columns will be unchanged, which is not ideal.
Any suggestions how can I achieve my goal.
Thank you
1 answer
Liftover gets you part of the way, yes. I assume you want to change the sequence as well.
Liftover to get the coordinates
fastaFromBED to get the sequence. This is in the bedtools package.
fastaFromBed -fi hg19.fa -bed seqs.bed -fo seqs.fa
You could filter the lines that contain ">" .
grep -v ">" seqs.fa > seqs.txt
Now you just have to stitch together your lifted over file and seqs.txt. I usually use python at this point.
I don't know about the specific differences between hg18 and hg19. In general, potentially, there might be a substitution, a deletion or insertion.
Log in to answer this question.