search

novoalab / EpiNano

Detection of RNA modifications from Oxford Nanopore direct RNA sequencing reads (Liu*, Begik* et al., Nature Comm 2019)
GNU General Public License v2.0
110 stars 31 forks source link

General Questions Regarding Using EpiNano #34

Closed vahidAK closed 4 years ago

vahidAK commented 4 years ago

Hi @Huanle , Thanks for updating EpiNano and making it more user-friendly. I am a bit confused about how to implement EpiNano. From my understanding, we extract per.site.var.slide features from fastq files and then we can use this file with your trained model to predict modifications? so we do not need fast5s at all? But in the previous version, you had feature extraction from Fast5 files, which extracts features from the event table in fast5s! what was that for? because in your new instructions I do not see any command that needs Fast5s as the input file!

Thanks, Vahid.

Huanle commented 4 years ago

@vahidAK sorry for getting back to you later. The trained models were derived without taking into account the current intensity values extracted from fast5 files. This is because we found that current intensity as an extra parameter does not improve prediction accuracy. So yes, you can proceed without extracting this information from fast5 files. However, this command line and the scripts for processing fast5 files are still available in case the users are interested in exploring it themselves. Hope this helps.

enovoa commented 4 years ago

Just to clarify, you can use release 1.0 to extract current intensities. Alternatively, you can use release 1.1 for faster prediction of modified sites.

vahidAK commented 4 years ago

Thanks a lot, @Huanle and @enovoa