search

nf-core / nascent

Nascent Transcription Processing Pipeline
https://nf-co.re/nascent
MIT License
18 stars 10 forks source link

Consider tsshmm #164

Open edmundmiller opened 1 day ago

edmundmiller commented 1 day ago

Description of feature

https://github.com/coregenomics/tsshmm @omsai any thoughts on using this over groHMM? I also noticed there's a fork of groHMM that I'm considering switching to.

omsai commented 1 day ago

Feel free to use the groHMM fork; I created that due to upstream being unresponsive to fixing bugs or adding any unit tests. For my work, I would keep and use all the models, because I think is important to look at various transcriptional regions of interest to the study in a genome browser and select the models that perform well for that dataset and research question. groHMM is a simpler 2-state model. TSSHMM has more states, which can be more useful for specimens with well-defined promoter regions or to make better use of 5'-capped RNA datasets.