Speed of step1 and model for Precision

[LipengKang]

Hi, professor

Thank you for developing this excellent software! It has been incredibly helpful for our work. However, I’ve encountered a couple of challenges that I hope you can provide guidance on:

1. Speed:

The initial processing step (Step 1) appears to be quite slow, even when working with datasets of relatively low coverage. Do you have any recommendations for optimizing or speeding up this step?

2. Low-Coverage Impact on Precision:

I am concerned about the potential for low coverage to affect the precision. Could you clarify if this is an inherent limitation of the software when working with sparse data?
Additionally, would it be possible for you to provide a pre-trained model for commonly used references like hg19 or other datasets?