search

morris-lab / CellOracle

This is the alpha version of the CellOracle package
Other
311 stars 56 forks source link

Selection of highly variable peaks before running cicero #89

Open PauBadiaM opened 2 years ago

PauBadiaM commented 2 years ago

Hi @KenjiKamimoto-wustl122,

Incredible package, its documentation and implementation are awesome!

In the preprocessing tutorial of scRN-seq, you recommend to do feature filtering and only select the top ~2-3k HVG to remove noisy genes and to reduce calculation time. I was wondering if the same idea could be applied and select the top N highly variable peaks before running cicero for the base GRN construction. In that notebook, there is no mention to it, but is this something you have played with? Or it would be something you would recommend? Thank you for your time.

fengweimin-maker commented 2 years ago

Hi @KenjiKamimoto-wustl122 ,

I also have the same question with PauBaiaM. Because It takes a lot of time to run cicero with all peaks for the base GRN construction.

Thanks !