Concatenate Sequence Fragments in Multiple Alignment Fasta

Question

[bhanratt]

I am using UCSC's multiz 100 species vertebrate multiple alignment fasta for hg19. It is refGene.exponAA.fa available here: http://hgdownload.cse.ucsc.edu/goldenpath/hg19/multiz46way/alignments/

The sequences seem to be broken up into fragments. For example the first sequence is:

>NM_152486.2_hg19_1_13 24 0 0 chr1:861322-861393+
MSKGILQVHPPICDCPGCRISSPV

In this example this is fragment 1 of 13. NM_152486.2_hg19_1_13 Further down there is 2_13, 3_13 etc.

I would like to concatenate all 13 fragments into 1 sequence for each refseq ID.

Is there existing software or a script that can perform this task?

Answer 1

[Chun-Jie Liu]

You may try sed -n '/NM_152486.2_hg19/{n;p}' refGene.fa |tr -d '\n'

[bhanratt]

Thanks for your response. I guess I didn't explain it very well. I need it to do it on all IDs and species and am just asking if anyone knows an existing method. Otherwise I can write one myself.

Thanks though!

[Chun-Jie Liu]

Use bash loop, grep and sed one line command can deal with this problem. I give the sed part.