mapping reads to consensus - O coverage regions

Question

[elina2410]

Hi, I performed reference based assembly of bacterial genome as followed: bwa-mem2 mapping reads to reference > marked duplicates with picard > variant calling with freebayes (-H 1) > filtered vcf based on DP>5 and VAF>0.5 with vcffilter > normalized indels with bcftools norm > bcftools consensus -H 1 with filtered and normalized vcf > mapping reads to consensus > polishing with pilon.

Then I calculated coverage samtools depth -d 0 -a based on reads mapped to consensus, and when I inspected coverage per base there are bases with 0 and other low coverage (1-5) bases. Why?

Best E.

[GenoMax]

If you are open to trying an alternate way then perhaps check out this answer by @Brian --> What is the correct procedure to generate a consensus bacterial sequence?

Did you inspect the alignment right after bwa step? Perhaps your genome is different from the reference used and thus the 0 coverage bases are there to begin with.

[elina2410]

Hi GenoMax , whole reference is uniformly covered by reads with small SNV/ins/del. Below is the region that is included into consensus but not covered by reads.