3
$\begingroup$

I am trying to use beagle 4.1 for phasing and IBD analyses. I was wondering if I need to run the analysis by chromosome or I can put them together in the same vcf file.

$\endgroup$
3
  • 1
    $\begingroup$ Well, I am new here, do you mind to give me some feedback on why the -1 on my question? Not trying to argue, just trying to improve. $\endgroup$
    – Max_IT
    Oct 31, 2018 at 13:36
  • 2
    $\begingroup$ I'm not sure who downvoted your question, but maybe because you don't have any info about what you have done so far or what you aim to achieve? In any case I think the question and the answer you need are clear enough, so +1. If you find out the answer to your question by some other means we would be happy if you submit an answer to your question. $\endgroup$
    – conchoecia
    Nov 1, 2018 at 15:49
  • 2
    $\begingroup$ Thank you for your explanation. I believe it should be clear enough for anyone familiar with the program. I ended up contacting one of the authors, I will post his answer here. $\endgroup$
    – Max_IT
    Nov 3, 2018 at 16:23

2 Answers 2

4
$\begingroup$

OK, I ended up contacting one of the authors of the program, who kindly answered to my question:

Either approach should work,

[i.e. running the analysis by chromosome vs joining the data together]

but I suggest dividing by chromosome because that approach has had more testing.

I hope this might help other users, I saw that several questions regarding Beagle went unanswered on biostars and seqanswers. I guess it is a pretty specialized topic.

Best,

Max

$\endgroup$
1
  • 1
    $\begingroup$ I don't see many questions about phasing and IBD detection on bioinformatcis SE. I think that's just because it's still a small site and there aren't that many people from that field on it yet. I wish there were more questions! $\endgroup$
    – winni2k
    Nov 25, 2018 at 14:05
1
$\begingroup$

I would also add that splitting your data into chromosomes allows you to run each chromosome in parallel (presuming you have some kind of server you can send each run to), vastly reducing the amount of time your analysis is going to take. I've not run all 22 chromosomes at once, but I can imagine it will take an enormous amount of time if you are using even a small number of markers and references.

$\endgroup$
1
  • 2
    $\begingroup$ This would be better served as a comment to the answer since it does not stand as a answer itself $\endgroup$
    – Bioathlete
    Nov 16, 2018 at 23:35

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.