search

VIB-PSB / MINI-AC

Motif-Informed Network Inference based on Accessible Chromatin (MINI-AC) is a method that combines accessible chromatin data from bulk or single-cell experiments with transcription factor binding site enrichment to learn gene regulatory networks in plants
Other
4 stars 0 forks source link

No enrichment found from snATAC seq data #25

Open BrunoGuillotin opened 3 months ago

BrunoGuillotin commented 3 months ago

Dear mini-Ac developers,

First thank you for developing such great tools. I used mini-Ex recently on my scRNAseq data with great success. I tried to run mini-Ac using data from snATAC seq (from 10x genomic multiome kit on Arabidopsis). While the pipeline ran correctly I got no enrichments in the output files whatsoever, for any of my 8 different cell types. I know mini-Ac was designed mostly to work on bulk ATACseq that is probably richer and less noisy than snATAC-seq but are there some parameters I could play with to improve the pipeline prediction on such datasets ? I get about from 10k to 2k specific peaks depending on the cell type.

Thanks in advance for your help, Bruno

hermandebeukelaer commented 2 months ago

Hi @BrunoGuillotin,

Sorry for the late reply. You can play with the P_val parameter to set a different threshold for the enrichment analysis. By default, the p-value threshold is 0.01 in locus-based mode, and 0.1 in genome-wide mode. Increasing the threshold should yield more enrichments in the output, but be aware that these are then statistically less significant.