Jennifer hi,
Thank you for the rapid answer. I will describe the problem I have, and maybe
you can see what is wrong:
1) I created a bed file called test.bed which contains the following text:
chr1 3585316 3777472 pos1
chr2 6585316 6585416 pos2
chr3 6585316 6585416 pos3
chr4 6585316 6585416 pos4
chr5 6585316 6585416 pos5
2) I placed test.bed in the directory where mm9ToRn4.over.chain is located, CD
to this directory, and started liftOver:
liftOver ./test.bed ./mm9ToRn4.over.chain ./test2.bed ./unMapped -errorHelp
-minMatch=0.95
3) This is the message I get:
Deleted in new:
None of sequence intersects with any alignment chain for the region
Partially deleted in new:
Sequence intersects with part of a single alignment chain in the region
Split in new:
Sequence partially intersects multiple chains in the region
Duplicated in new:
Sequence completely intersects multiple chains in the region
Boundary problem:
Missing start or end base in an exon
So:
I do not get any output.
I assume test2.bed and unMapped are the output files, but they were not created
by liftOver (should I create them first?). In addition what is "Missing start
or end base in an exon" (I used a genomic data, not genes)
It should be noted that if I use the same bed file as an input for the online
lifOver (old genome mouse, new genome human), I get some results:
conversions:
chr8 56090155 56244617 pos1
chr10 11299218 11299320 pos2
chr7 88976276 88976377 pos5
failures:
#Deleted in new
chr3 6585316 6585416 pos3
#Deleted in new
chr4 6585316 6585416 pos4
--- On Wed, 5/27/09, Jennifer Jackson <[email protected]> wrote:
> From: Jennifer Jackson <[email protected]>
> Subject: Re: [Genome] liftover parametrs
> To: "Fungazid" <[email protected]>
> Cc: [email protected]
> Date: Wednesday, May 27, 2009, 1:04 AM
> Hello Avi,
>
> If you type in the command at the prompt, a list of options
> will appear along with usage. The over.chain files are in
> Downloads, ftp directory
> goldenPath/<from_species>/liftOver/.
>
> For default genomic examples, there are two classes: same
> species lifts and cross-species lifts. Both can be seen in
> the web version [UCSC Browser home page -> Utilities
> -> Batch Coordinate Conversion (liftOver)]
> http://genome.ucsc.edu/cgi-bin/hgLiftOver
>
> The most basic option for genomic would be to set -minmatch
> = 0.95 for same-species and = 0.1 for cross-species and
> leave the rest as default.
>
> For any failures, you can examine the reasons and play
> around with the parameters, such as setting multiple = Y.
> But beware that this can produce significant output if
> minmatch is low and chain and hit size are set >= 0. It
> will probably take a few attempts to tune these in various
> combinations so that the output is useful for your
> particular needs.
>
> The other options come into play for gene, transcripts and
> other annotation types that are lifted. Again, the program
> runs pretty fast even on large datasets, so a cyclic
> test-and-analyze approach is probably the best way to find
> your own best parameters to create useful lifted data.
>
> FAQ in case you have not seen it:
> http://genome.ucsc.edu/FAQ/FAQdownloads#download28
>
> Good luck and please write back if you need more technical
> help,
> Jennifer Jackson
> UCSC Genome Bioinformatics Group
>
> Fungazid wrote:
> > hello genome browser staff,
> >
> > I would like to map species A genomic coordinates to
> species B genomic coordinates (e.g. mouse -> human). To
> my knowledge liftover can do it (I also tried to workaround
> by blating the query region with gfserver/gfclient, but this
> is too slow for large tables). I have Liftover installed on
> Linux, but I didn't find any documentation about input
> parameters and examples. Maybe you can help, or direct me to
> a proper documentation.
> >
> > Thanks, (also you for your previous help)
> > Avi
> >
> >
> >
> _______________________________________________
> > Genome maillist - [email protected]
> > https://lists.soe.ucsc.edu/mailman/listinfo/genome
> >
_______________________________________________
Genome maillist - [email protected]
https://lists.soe.ucsc.edu/mailman/listinfo/genome
