Majority of SNPs have allele frequency = 0.000000

[ryanskiba]

Hi,

I'm using bcftools to call and filter markers for a haploid biparental population of 118 individuals. I have Illumina sequencing for all progeny. After indexing my reference genome (one of the parents of the biparental population), converting my .fq files to .sam to sorted .bam files, I called markers using:
bcftools mpileup -Ou -f genomic.fasta *.bam | bcftools call -mv -Ob --ploidy 1 --threads 4 -o calls1.bcf

and convert to .vcf using: bcftools view -Ov calls1.bcf > test.vcf

I then run bcftools stats calls1.bcf > stats.txt.

I don't understand why, out of the 3.2 million SNPs in my .bcf file, roughly 1.9 million have an allele frequency of 0.000000. Additionally when I use grep -c "AC=0" calls1.vcf, it gives me a result of 0. So I'm not sure how to look at these markers that are supposedly present at such low frequency. Nor do I understand why markers would be called if they truly aren't present in the population.

Here's a portion of the data from bcftools stats:

AF, Stats by non-reference allele frequency:

AF [2]id [3]allele frequency [4]number of SNPs [5]number of transitions [6]number of transversions [7]number of indels [8]repeat-consistent [9]repeat-inconsistent [10]not applicable

AF 0 0.000000 1925317 1318327 606990 271 0 0 271
AF 0 0.008475 25134 23556 1578 34 0 0 34
AF 0 0.016949 86417 79970 6447 110 0 0 110
AF 0 0.025424 82610 76945 5665 115 0 0 115

I can get rid of these by filtering with a minor allele frequency, but I'd like to know why they were present in the first place.

Thanks!