How to compare two vcf files of two cultivars?

shis

Hi,
I am looking for a program that could help to compare two vcf files of two cultivars. I have three vcf files (vcf1, vcf2, vcf3) of SNPs of three different cultivars. The SNPs were called using GATK's UnifiedGenotyper. After stringent filtering with GATK's hard filters, I found approx. 6 millions SNPs in these three cultivars. Now I would like to compare two vcf files of SNPs of any two cultivars (e.g., the combination can be: vcf1/vcf2, vcf1/vcf3 and vcf2/vcf3) to identify common and specific SNPs, and also like to generate venn diagrams between the combinations.

Can anyone suggest me - which tool I can use to identify common and specific SNPs between two cultivars? I do appreciate your help.

Sheila

@shis
Hi,

Have a look at this page which should help.

-Sheila

Sheila

@shis
Hi,

Have a look at this page which should help.

-Sheila

shis

@Sheila
Thank you so much for the link. I got the intersection from the combined vcf file of the two vcf files using the following command:

combine the data

java -Xmx2g -jar $GATK -T CombineVariants -R IRGSP.fasta -V cultivar1.vcf -V cultivar2.vcf -o union.vcf

select the intersection

java -Xmx2g -jar $GATK -T SelectVariants -R IRGSP.fasta -V union.vcf -select 'set == "Intersection";' -o intersect.vcf

Now I would like to separate SNPs that are not in the intersect area. How can I get SNPs that are different from each other?

Sheila

@shis
Hi,

Right now, you have all the sites where the two cultivars are variant. You can simply invert the selection criteria in SelectVariants using --invertselect. That will choose the sites not in set==intersection.

-Sheila

Geraldine_VdAuwera

You can also select individual sets by name to isolate them to separate files, if that is of interest. Check out the workshop tutorials for more details on how to do this.

shis

Hi Sheila,
According to your suggestions, I used the following command to get variants from the two cultivars where the variant sites is not 'set==intersection':
$ java -Xmx16g -jar $GATK -T SelectVariants -R ./ref_genome/IRGSP.fasta -V cultivar1_cultivar2_union_final.vcf -select 'set == "Intersection";' -o cultivar1_cultivar2_variant_final.vcf -invertSelect

Now my question is: are these SNPs found between two cultivars with respect to reference genome or without reference genome?
Thanks for the help.
/Shis

@Sheila said:
@shis
Hi,

Right now, you have all the sites where the two cultivars are variant. You can simply invert the selection criteria in SelectVariants using --invertselect. That will choose the sites not in set==intersection.
-Sheila

Sheila

@shis
Hi,

You now have the sites that are either variant in one set or variant in the other. Those sites will always be in respect to the reference, but will have different genotypes from each other. I hope this helps.

-Sheila

shis

@Sheila, many thanks for the answer.