search

PengNi / deepsignal-plant

Detecting methylation using signal-level features from Nanopore sequencing reads of plants
GNU General Public License v3.0
54 stars 12 forks source link

input signal data of deepsignal #39

Open sengelen opened 9 months ago

sengelen commented 9 months ago

Hi Peng

I have many nanopore sequencing runs from an arabidopsis strain with fail and pass data. I was wondering if it is better take all the data, only pass data or only data from the same sequencing run ? And which coverage do you recommend ?

Best Stefan

PengNi commented 9 months ago

Hi Stefan, for arabidopsis I'd recommend 40-100x coverage of reads. I recommend to use PASS reads only if you have enough PASS data. I am not sure if there is any significant batch effect if you use data from different runs.

Best, Peng

sengelen commented 9 months ago

Thank you

Best, Stefan