Samtools / Bcftools Missing Random Obvious Single Base Indels

Question

[Torst]

I am using samtools to call variants in a haploid genome (yes I know it is designed for diploid). It finds SNPs easily, and most of the indels, but there are a few indels it can NOT seem to find no matter what parameters I give it. I use bwa mem for the alignment of MiSeq PE 150 reads, and I get about 150x coverage, all good. Using tview etc, there is clearly a G deletion at position 600, but bcftools will not seem to call it, despite calling some nearby (~100bp before). All the mapping and base qualities seem to be fine.

This is the command I am using:

samtools mpileup -B -g -u -f ref.fa aln.bam | bcftools view -v -

It happens whether I use the -B or not.

Any help appreciated.

Below is the INDEL position pileup +/- 2 bases:

Contig_Plasmid  598     C       185     ....,,,..,,,.,,,,,,..........,...,.,,,....,,.$,$,.,....,.,,...,,......,,,...,,,,,,,,,,,,,,....,...,,,.,.....,.,,.,....................,,,,,,,,.....,.,.,..,,,,,,
,,,,,,,.,,.....,,,,,,..,..,     BHHHCBFHHGGGHGGGGGGHHGHHGDHHHGHHHGHHHHFHHHHHHAHHHHHHGGGHHFHHIGHFHHHHHHHHHHHHGHHHHHHHHHEHD5FH2HH?GHHHGIHHIHHHFHHGGHHHBHHHHGGHHHHHHHHHFHHHHGHHFGHHHHHHHHHHHGHGHHHHHHH?HHBH
FFFFFHHHHHHB@HCDH
Contig_Plasmid  599     A       183     ....,$,$,..,,,.,,,,,,..........,...,.,,,....,,,.,....,.,,...,,......,,,...,,,,,,,,,,,,,,....,...,,,.,.....,.,,.,....................,,,,,,,,.....,.,.,..,,,,,,,,
,,,,,.,,.....,,,,,,..,..,       EGHHCBFHHGGFHFGGGGGHHHFHHFHHHGHHHFHHHHHHHHHHHHHHHHFHHHHHHHHFHGHHHFHIHHHHHHFHHHHGHHHHAHGDHH2HH1GHHHHHHHHHHHFHHHHHHHGHHHHHHHHHHHHHHHFHFHHHHH5HHGHHHHHHHHHHHHHHHHGII4HHFHFF
FFFHHHHHHB?HCDH
Contig_Plasmid  600     T       181     .$.$.$.$,.-1G.-1G,-1g,-1g,-1g.-1G,-1g,-1g,-1g,-1g,-1g,-1g.-1G.-1G.-1G.-1G.-1G.-1G.-1G.-1G.-1G.-1G,-1g.-1G.-1G.-1G,-1g.-1G,-1g,-1g,-1g.-1G.-1G.-1G.-1G,-1g,-1g,-1
g.-1G,-1g.-1G.-1G.-1G.-1G,-1g.-1G,-1g,-1g.-1G.-1G.-1G,-1g,-1g.-1G.-1G.-1G.-1G.-1G.-1G,-1g,-1g,-1g.-1G.-1G.-1G,-1g,-1g,-1g,-1g,-1g,-1g,-1g,-1g,-1g,-1g,-1g,-1g,-1g,-1g.-1G.-1G.-1G.-1G,-1g.-1G.-1G.-1G,-1
g,-1g,-1g.-1G,-1g.-1G.-1G.-1G.-1G.-1G,-1g.-1G,-1g,-1g.-1G,-1g.-1G.-1G.-1G.-1G.-1G.-1G.-1G.-1G.-1G.-1G.-1G.-1G.-1G.-1G.-1G.-1G.-1G.-1G.-1G.-1G,-1g,-1g,-1g,-1g,-1g,-1g,-1g,-1g.-1G.-1G.-1G.-1G.-1G,-1g.-1
G,-1g.-1G,-1g.-1G.-1G,-1g,-1g,-1g,-1g,-1g,-1g,-1g,-1g,-1g,-1g,-1g,-1g,-1g.-1G,-1g,-1g.-1G.-1G.-1G.-1G.-1G,-1g,-1g,-1g,-1g,-1g,-1g.-1G.-1G,-1g.-1G.-1G,-1g     4FHEBHHFFFHFGFGGGHHHHHHFHHHGHHHFHHGHHHHHHHHHHHHHHGHHHFHHIFHGFHHHHHIHHHGHFHHHHHHHGH5H55HHDHGFGHHHFHHHHHHHFHHHHGHHFHHHHHHHHHFHGHHHHHFHHHIGGHHGHHHHHHHHHHHGHHEH2HH4HGHHFFFFFHHHHHHBAHDDH
Contig_Plasmid  601     G       179     ,*********************************************************************************************************************************************************************************^]t   BHHFFFHFGGGGGHHGHHHHHHHGGHHHFHHHHHHHHHHHHHHHHHHHHHFHHHFHBHHHHGIIHHHHIFHHHHHGHHHHHFDHHDHHFGHHHHHHHHHGHDHHGFHHGBHHHHHGHHHHHHHHHGHFHHHIHHHHHHHHHHHHHHHHHHHHHHHH4HGHHGGGGGHHHGHHFFHCDHB
Contig_Plasmid  602     G       179     ,..,,,.,,,,,,..........,....,.,,,....,,,.,....,.,,...,,......,,,...,,,,,,,,,,,,,,....,...,,,.,.....,.,,.,....................,,,,,,,,.....,.,.,..,,,,,,,,,,,,,.,,.....,,,,,,..,..,,     BHHFFFHFGGGGGHHGHHHHHHHGGHHHFHHHHHHHHHHHHHHHHHHHHHFHHHFHBHHHHGIIHHHHIFHHHHHGHHHHHFDHHDHHFGHHHHHHHHHGHDHHGFHHGBHHHHHGHHHHHHHHHGHFHHHIHHHHHHHHHHHHHHHHHHHHHHHH4HGHHGGGGGHHHGHHFFHCDHH

[Istvan Albert]

I hope this gets answered, we've occasionally noticed similar situations and could not explain it, left us scratching our heads that some other cutoff/threshold might have been triggered

[Chris Fields]

@Torst, have you tried any other callers? Just curious if they do call this...

[Torst]

I also found this odd behaviour too, which I filed as an Issue on the samtools github site: https://github.com/samtools/samtools/issues/79