search

WGLab / RepeatHMM

a hidden Markov model to infer simple repeats from genome sequences
Other
34 stars 14 forks source link

MinION data #26

Closed woodoo46 closed 4 years ago

woodoo46 commented 5 years ago

Hi there,

Do we need to do some special processing for MinION data?

Thanks.

liuqianhn commented 5 years ago

Hi @woodoo46 , you can set `--hmm_insert_rate, --hmm_del_rate and --hmm_sub_rate for insertion error rate, deletion error rate and mismatch error rates for specific sequencing platform, which might improve the performance but not fully tested.

woodoo46 commented 5 years ago

Do you have recommended settings for MinION data? Thanks.

liuqianhn commented 5 years ago

I have no optimized settings for MinION data. But you can set insertion rate as 0.03, deletion rate as 0.05 and mismatch as 0.06. But there is no guarantee that the parameters are optimal.

woodoo46 commented 5 years ago

These settings are not data error rate, right? They are for hmm parameters?

liuqianhn commented 5 years ago

@woodoo46 , they would be data error rates. So, it would be better if you could provide optimal error rate based on your data, although usually it does not affect the results significantly.