liftover genome build from GRCh 37 to 38

[MAOMAO]

Hi All,

I have a question for lifting genome build from GRCh 37.p13  to GRCh38.p11.  I am trying to use UCSC website and NCBI remapping service to load my .bed file to convert, but failed. I saw lots of example are just put few lines of Chr starting position end position. 

Any suggestions would be appreciated.

MAOMAO

[Christopher Chang]

https://www.cog-genomics.org/plink/1.9/formats#bed

See the boldface note.  Unfortunately, there are *two* common uses of the ".bed" file extension in genomics.

It is still possible to lift over a .bed+.bim+.fam fileset from one genome build to another, but it requires substantially more work.  Some variants become unmapped, some variants change direction...

[Christopher Chang]

(...and if you're dealing with sequencing data, the result of this type of liftover would not be as good as you'd get from realigning the original reads to the new reference, and then calling variants on the realigned reads.)

[Ying Zhao]

Thanks so much Christopher for your clarification. 

My genotyping data is based on GRCh37, and I used 1kG genotyping (GRCh38) as reference. Which way would be easier and faster if I want to keep them in the same coordiniate?

MAOMAO

--
You received this message because you are subscribed to the Google Groups "plink2-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to plink2-users...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/plink2-users/08770c23-8d89-4cd5-85a9-4008b3f8bb10%40googlegroups.com.

[Shicheng Guo]

Hi Maomao and Christopher, 

Any best practice for plink1.9 to liftOver from hg19 to hg38 for BIM/FAM/BED files comes out? 

Thanks.

Shicheng

On Mon, Apr 13, 2020 at 4:21 PM Christopher Chang <chrch...@gmail.com> wrote:

--

[Christopher Chang]

There are a bunch of tools that can perform this operation on a VCF; pick one you're comfortable with.

[Chris Chang]

If you want to try to do this manually, the instructions at https://www.protocols.io/view/genotyping-chip-data-lift-over-to-reference-genome-xbhfij6 look reasonable to me.

On Wed, Jul 28, 2021 at 8:46 PM Shicheng Guo <shihch...@gmail.com> wrote:

VCF based liftOver definitely is the most secure approach, however, it requires a huge hard disk requirement for biobank-level data. I am wondering whether plink-bim-based liftOver can also achieve the same accuracy.  Thanks. 

To view this discussion on the web visit https://groups.google.com/d/msgid/plink2-users/ba15e6c8-665d-4f76-9835-808f410a7ddbn%40googlegroups.com.