Questions related to read file, sample, and VCF

[nrayamaj]

Hi there,

How does the software know that the given read file belongs to the particular sample if the VCF has multiple samples?

What we need to do to ensure that the read file corresponds to the correct sample for genotyping?

Can we parallelize our task if we have multiple samples and read files?

Regards,
Niraj

[clemaitre]

Hi Niraj,

For the moment, SVJedi-graph can genotype only one sample at a time. The genotyped sample is the one corresponding to the read file given as input. SVJedi-graph ignores the information of genotypes already present in the input VCF. If the input VCF contains one or several sample columns with genotypes, the output VCF will not contain any of these columns and instead there will be only one genotype column, with the genotypes that have been estimated by SVJedi-graph with the input read file.

If you have several samples to genotype, you will have to run SVJedi-graph for each sample separately and then combine all the output VCFs. We are planning to implement a multi-sample mode, automatizing this process, but we have not found the time yet to do it...

Regards,
Claire