Some questions about understanding the output of qcSTR

[YiweiNiu]

Hi,

Thanks for developing this useful tool. I have two questions about understanding the output of qcSTR.

1. In the doc of qcSTR (https://trtools.readthedocs.io/en/latest/source/qcSTR.html):
   

Can be used to visualize if there is a strong bias toward calling deletions vs. insertions compared to the reference, which might indicate a problem.

How to interpret this plot? For example, would it indicate a low quality of calls if a skew toward expansion was observed?

2. A related question:
   

It is expected that the mean difference should be around 0 for most settings. When this value starts to deviate from 0, e.g. for very long repeats, it could indicate a drop in call quality.

I am confused about this "around 0" expectation. What's the assumption behind this?

I am new to this field. Sorry if my questions are too naive.

Thanks in advance.

Bests,
Yiwei