search

ijuric / MAPS

18 stars 11 forks source link

Few interactions can be detected by MAPS #8

Open zhuyezhang opened 5 years ago

zhuyezhang commented 5 years ago

Hi, I tried to run MAPS on ESC H3K27ac PLAC-Seq data from Renbing lab(GSE86150) currently. I downsampled the data to 100M reads and ran MAPS at 5kb resolution. I could only get about 400 significant interactions and could not get the interactions showed in their paper. Did anyone ever run MAPS on this data? BTW, how many reads should be proper for MAPS analysis? Thanks for your help in advance!

Yezhang

armenabnousi commented 3 years ago

Hi Yezhang,

I'm not aware if anyone has ran MAPS using this dataset. Generally I think 100M is too little, I'd recommend using 10Kb for datasets from 200M to 450-500M, and 5KB for >500M. Sorry about the delayed response.

Thanks