VEP annotation

[pontikos]

Use build 38.

Add ExAC annotation allele frequencies:
/cluster/project8/vyp/AdamLevine/annotations/ExAC/0.3/ExAC.r0.3.sites.vep.vcf.gz
Have a look at this script to prepare allele frequency from INFO:
/cluster/project8/vyp/AdamLevine/annotations/esp/prepare_esp.sh

Add 1kg allele freq.

Add CADD scores

[pontikos]

I've committed a couple of scripts in annotation to prepare the ExAC custom annotation for VEP:
a668095
and
65c5d97

[pontikos]

I've written a generic liftover script to prepare the custom annotation for VEP.
The data needs to be split by chromosome in order for it to work.
Note that it doesn't deal with positions moved to a different chromosome or with alternative sequences.
These just get dropped by filtering the bed file on the chromosome name.

defa3a5

[APLevine]

Sounds good.

Adam

Adam P. Levine
On 25 Jan 2015 23:42, "Nikolas Pontikos" notifications@github.com wrote:

I've written a generic liftover script to prepare the custom annotation
for VEP.
The data needs to be split by chromosome in order for it to work.
Note that it doesn't deal with positions moved to a different chromosome
or with alternative sequences.
These just get dropped by filtering the bed file on the chromosome name.

defa3a5
defa3a5

—
Reply to this email directly or view it on GitHub
#21 (comment).

[pontikos]

Thanks.

I'm moving the annotations to /cluster/project8/IBDAJE/VEP_custom_annotations
Initially I had moved them to /goon2/scratch2/vyp-scratch2/annotation but I found writing to that location to be quite unreliable using the SGE (sometimes output files were empty).

I have updated the custom annotations in run_VEP.sh 2d50034 to read from a genome build dependent location

[APLevine]

Excellent. So now the post-VEP script just has to be finished. Let me know
if you want to discuss what this is going to do.

Adam

Adam P. Levine
On 25 Jan 2015 23:52, "Nikolas Pontikos" notifications@github.com wrote:

Thanks.

I'm moving the annotations to
/cluster/project8/IBDAJE/VEP_custom_annotations
Initially I had moved them to /goon2/scratch2/vyp-scratch2/annotation but
I found writing to that location to be quite unreliable using the SGE
(sometimes output files were empty).

I have updated the custom annotations in run_VEP.sh 2d50034
2d50034
to read from a genome build dependent location

—
Reply to this email directly or view it on GitHub
#21 (comment).

[pontikos]

Will do.
I'm noticing some strange things occurring on the filesystem though: sometimes the output of bgzip is empty.
I don't think it's due to a bug in my script but instead probably because of some NFS lag.
I'm looking into it.

[pontikos]

Ok I think I know what it is: I had a chr${ch}*.vcf.gz instead of chr${ch}_*.vcf.gz so there must have been some concurrency issues because the same file would be matched twice. Correcting that and running again.
I've committed the fix 60aae87

[pontikos]

See #25 , the liftover of the annoations form 37 to 38 is not trivial because certain regions (especially around the centromeres) have changed significantly.

[APLevine]

I wonder how the ESP annotation in b38 works with the VEP?

Adam

Adam P. Levine
On 29 Jan 2015 15:08, "Nikolas Pontikos" notifications@github.com wrote:

See #25 #25 , the liftover
of the annoations form 37 to 38 is not trivial because certain regions
(especially around the centromeres) have changed significantly.

—
Reply to this email directly or view it on GitHub
#21 (comment).

[pontikos]

Good point, yes I am not sure how the built-in ESP annotation works compared to the custom one which we liftover. I will check.