LiftOver Question

[Doherty, Aoife]

I apologise if this is not the appropriate email for my question, please let me know if it is better directed elsewhere. 

I want to transform a set of SNPs from the canFam2 to canFam3 positions on chromosomes. 

I tried to use the web version of liftOver, but when I load the canFam bed file and select the relevant parameters and click submit, the page just refreshes (in both safari and chrome).

So then I downloaded the linux version, and ran the command:

./liftOver File.bed canFam2ToCanFam3.over.chain newFile unMapped

This results in the error:

Reading liftover chains

Mapping coordinates

Data format error: expecting at least 3 fields in BED file (File.bed)

I'm just using a standard bed file that I generated from plink, I can't see anything unusual about it. Am I using the wrong type of file? (e.g. the command that would have given me the bed file would have been: plink --file File --make-bed --out File --noweb --dog).

Many thanks

Aoife

[Hiram Clawson]

Good Morning Aoife:

Can you please show a couple of the lines from the File.bed file ?

--Hiram

[Cath Tyner]

Hello Aoife,

Thank you for posting your question regarding the use of liftOver with a PLINK BED file.

Note that there are differences between a PLINK BED (binary PED) file and a UCSC BED (Browser Extensible Data) file. You can review the UCSC BED file format and compare this to your PLINK BED format.

I did find two resources which may help you:

1. crossMap, a program for conversion of genome coordinates between assemblies (can lift VCF directly).

2. https://www.researchgate.net/post/How_can_I_create_a_UCSC_BED_file_from_plink_data

If your BED file is in the correct UCSC BED format and you are still having problems, please respond to the group for further assistance. In doing so, it would be helpful for us to see a few lines from your file so that we can help verify the format.

Thank you again for your inquiry and for using the UCSC Genome Browser. If you have any further questions, please reply to gen...@soe.ucsc.edu. All messages sent to that address are archived on a publicly-accessible forum. If your question includes sensitive data, you may send it instead to genom...@soe.ucsc.edu.

Enjoy,
Cath
. . .

Cath Tyner

UC Santa Cruz Genomics Institute

UCSC Genome Browser: Public Help Forum, Suggestions, Contact

--

---
You received this message because you are subscribed to the Google Groups "UCSC Genome Browser discussion list" group.
To unsubscribe from this group and stop receiving emails from it, send an email to genome+un...@soe.ucsc.edu.