Mapping Probe Coordinates To A Genome Build?

Question

[Darren J. Fitzpatrick]

Hi,

I have a collection of expression probe sequences, their chromosomes and their coordinates. I wish to identify which genome build my coordinates come from. Is there a simply way to do this?

Thanks, D.

[Aaronquinlan]

You say you know the sequence of the probe, as well as the coordinates? If so, just paste a few of your coordinates into UCSC and click the "DNA" button up top. If it matches your probe sequence, you have the right build. Else, try the other build.

[Aaronquinlan]

you know the species, right?

[Darren J. Fitzpatrick]

The species is human - I just need to know if they are from NCBI 36 or GRCh37.

[Vikas Bansal]

Ahh. If you know the probe sequences, then aaron is right. I thought you have only chromosome, start and end position.

[Darren J. Fitzpatrick]

@aaronQuinlan - that worked a treat! I am knee deep in UCSC documentation. I have never used this resource before. It's super! If you wish to submit your comment as an answer, i'll mark it correct.

Answer 1

[Vikas Bansal]

Assuming that you are talking about human genome. As there is difference between chromosome lengths of hg17, hg18 and hg19, you can check your BED file (chromosome, start and stop) for each chromosome. Example- chromosome1 in hg18 has 247,249,719 bp and hg19 has 249,250,621 bp. If you have any coordinate with chr1 248,250,621 249,150,621 then you can say the probes belong to hg19. You have to do this for all chromosomes.

But the problem is, if your probes are from hg19 but all the coordinates lies in hg18 also.

You can give it a try, and if it is from hg19 and you got some coordinate as I showed in example, your problem will be solved. I really want to know if we have any good solution for this.

[Gustavo]

Ah, but your solution can be extended to test for reference gaps as well... if something overlaps a gap, chances are you're looking at the wrong ref. :)