Discrepancy between GATK DepthOfCoverage and samtools mpileup - depth values

astrandastrand New YorkMember


I was comparing the output from GATK DepthOfCoverage and samtools mpileup and noticed that for chromosome 1, for instance, the depth values differ per individual:

GATK DepthOfCoverage sample (4th column is for individual of interest):

1:10001 1056 39.11 43
1:10002 1973 73.07 91
1:10003 2728 101.04 120

samtools mpileup (the last field is for the individual of interest):

1 10001 . T 0 . DP=298;I16=246,10,0,0,8031,252873,0,0,266,4956,0,0,216,3484,0,0;QS=8,0;MQSB=0.882047;MQ0F=0.855705 PL:DP:DV:SP:DPR 0,120,16:40:0:0:40,0
1 10002 . A 0 . DP=605;I16=492,13,0,0,15628,487360,0,0,583,11687,0,0,480,4056,0,0;QS=8,0;MQSB=0.82856;MQ0F=0.86281 PL:DP:DV:SP:DPR 0,244,26:81:0:0:81,0
1 10003 . A C, 0 . DP=862;I16=727,19,1,1,22640,694394,40,808,1252,27124,0,0,1102,6936,4,8;QS=7.98193,0.018075,0;VDB=0.02;SGB=-2.77779;RPB=0.485255;MQB=0.873995;MQSB=0.836835;BQB=0.0589812;MQ0F=0.845708 PL:DP:DV:SP:DPR 0,255,48,255,51,48:112:1:0:111,1,0

Any thoughts as to what could be happening?



  • SheilaSheila Broad InstituteMember, Broadie, Moderator admin

    Hi Alva,

    DepthOfCoverage does filter out some reads. However, it looks like samtools is reporting less reads than DoC. Maybe samtools applies some different/extra filters? To debug, you can try looking at a small region with a few reads in IGV to see which reads are being counted by each tool.


Sign In or Register to comment.