Incongruities at program results

[dukelheit]

Hello, im using SNPmatch in a VCF file where i have 7 replicates of 3 different plants (21 samples in total), i made the database with them and then i splitted in 21 separated VCFs to run the program with each one, but when i ran the program 21 times and made a heatmap to see the "probability of match" results, i see incongruents. For example, see the subsets RGS vs REED, subset 1 is showing more simmilarities than subset 2, how must i interpret this?

[rbpisupati]

Hi @dukelheit! can you please explain a bit more on what functions are you using? does you VCF file only contains biallelic sites? are you using any other filtering while generating VCF?

[dukelheit]

Hi again.
Sure, here i'll let the code which i has used to process my vcf:

"FILTER THE VCF (BIALLELIC INCLUSIVE)
bcftools view -i 'INFO/MQ>40 && INFO/FS<60 && INFO/QD>2 && INFO/SOR<3.0 && INFO/MQRankSum>-12.5 && INFO/ReadPosRankSum>-8.0 && %QUAL>=30' --max-alleles 2 --exclude-types indels -Ov named.vcf.gz > named.vcf

"SPLIT THE VCF FILE"
for sample in bcftools query -l named.vcf; do java -jar /home/franco/BinsPortables/gatk-4.1.5.0/gatk-package-4.1.5.0-local.jar SelectVariants -V named.vcf -O ${sample}.vcf -sn $sample; done

"MAKE THE DATABASE"
snpmatch makedb -i named.vcf -o DBs/db

"RUN SNPmatch"
for sample in $(ls | grep .vcf$); do snpmatch inbred -d DBs/db.hdf5 -e DBs/db.acc.hdf5 -i $sample -o Op/output_snpmatch_${sample}; done

Finally i just plot the 21 x 21 "probability of match" results
Thanks in advance
Franco

[rbpisupati]

Thank you! I still do not see why you couldn't see why you are seeing such results. We use the same functions to identify identical strains if any within our datasets. I am not sure if this is a bug in the program, or with data you are using.