There appears to be a bug with the reference base reported by mpileup -aa in certain situations.
With the newest samtools (1.22), the following:
tr '|' '\t' > test.sam << EOF
@SQ|SN:S1|LN:2
@SQ|SN:S2|LN:1
@SQ|SN:S3|LN:1
X1|0|S1|1||1M|*|||A|*
X2|0|S2|1||1M|*|||A|*
X3|0|S2|1||1M|*|||A|*
X4|0|S3|1||1M|*|||A|*
EOF
cat > test.fasta << EOF
>S1
TG
>S2
G
>S3
G
EOF
samtools mpileup -aa -f test.fasta test.sam
Produces output that contains a non-ascii character in place of the reference base (on the second line:
[mpileup] 1 samples in 1 input files
S1 1 T 1 ^!A$ ~
S1 2 0 * *
S2 1 G 2 ^!A$^!A$ ~~
S3 1 G 1 ^!A$ ~
Prior to #2019 the reference base would be N, which while wrong did not break downstream pipelines as much.
There appears to be a bug with the reference base reported by
mpileup -aain certain situations.With the newest samtools (1.22), the following:
Produces output that contains a non-ascii character in place of the reference base (on the second line:
Prior to #2019 the reference base would be
N, which while wrong did not break downstream pipelines as much.