huangyh09 / brie

BRIE: Bayesian Regression for Isoform Estimate in Single Cells
https://brie.readthedocs.io
Apache License 2.0
41 stars 15 forks source link

What's the difference between mouse_features.filtered.csv.gz and mouse_features.lenient.csv.gz, and which one to use? #17

Closed gangcai closed 5 years ago

gangcai commented 6 years ago

Hi Yuanhua, For the parameter "-f" in brie, which feature file should I use? The lenient or filtered? Is "-f" parameter optional or must needed?

Thanks, Gangcai

RuyBlast commented 6 years ago

Hi Gangcai, I worked on BRIE, I do not know where you found your two feature files, but I can tell you that -f is a needed argument for BRIE. You can generate your feature file thanks to brie-factor from briekit. But if you want to try quickly, I think from their names that you could use one or the other of your feature files (I would use the filtered one). Best, Milan

gangcai commented 6 years ago

Hi Milan, Thanks for your reply. Those two files were generated by briekit-factor according to this script provided by Brie https://github.com/huangyh09/briekit/blob/master/example/anno_mouse.sh. The names indicated in the script are mouse_factors.SE.filtered.csv.gz and mouse_factors.SE.lenient.csv.gz respectively. Do you have any suggestion about in which scenario should I use filtered and when to use lenient?

Best, Gangcai

huangyh09 commented 6 years ago

Hi Gangcai,

These two files are according to the Splicing event annotation: SE.filtered.gff3.gz and SE.lenient.gff3.gz. The former is stricter while the latter is more lenient on the length of alternative exon and the distance between TSS or TTS. Normally, I would recommand using SE.filtered.gff3.gz, but if it doesn't include the events you are interested in, then you could use the lenient one or add it manually.

Best, Yuanhua