# Why does samtools mpileup sometimes include ref bases (other than ',' or '.')?

This is my first post here. I can think of no way of giving an easily reproducible example, as per the stack ethos. SO apologies in advance and any feedback on question format appreciated.

I have called samtools mpileup (Version: 1.9 (using htslib 1.9)) on some oxford nanopore data (mapped to GRCh38).

I am seeing some instances like below (grouped with pythons Counter), here reference is T at this position:

 Counter({',': 34, '.': 33, 'c': 9, 'g': 6, 'a': 4, 't': 2, '-:ACTC': 1, '-:ACTCGG': 1, '-:5': 1, '\$': 1, '*': 1})


Why is 't': 2' in there?

flags used: -A -B -Q 1 -R -x

• Hi Liam, can you show the exact mpileup for this position? Because I see that you tried to parse this into python and also converted the cases (no upper cases). Feb 27, 2020 at 14:10
• This way it's easier to see exactly was called. Feb 27, 2020 at 14:12
• Thanks @StupidWolf . I stupidly didn't keep the position of the above and it was just one example of a big list. So I'll edit above and give more examples and more details. I appreciate the help. At this point I'm thinking mpileup is a bit buggy, hoping that its resolvable. Feb 28, 2020 at 1:37