search

FredHutch / SEACR

SEACR: Sparse Enrichment Analysis for CUT&RUN
GNU General Public License v2.0
99 stars 43 forks source link

SEACR run failed for control mode #86

Open sfpacman opened 2 years ago

sfpacman commented 2 years ago

I wasn't unable to run SEACR (top 0.01 run just fine) when I use control bedgraphs with the following command: bash SEACR_1.3.sh $sample_bed $control_bed non stringent ${sample}_seacr_ctrl.peak

Using stringent threshold
Creating experimental AUC file: Fri Jun  3 03:17:29 PDT 2022
Creating control AUC file: Fri Jun  3 03:18:20 PDT 2022
Calculating optimal AUC threshold: Fri Jun  3 03:18:30 PDT 2022
Calculating threshold using non-normalized control: Fri Jun  3 03:18:30 PDT 2022
Creating thresholded feature file: Fri Jun  3 03:19:41 PDT 2022
Empirical false discovery rate = 7.62799780318346e-06
Merging nearby features and eliminating control-enriched features: Fri Jun  3 03:19:42 PDT 2022
Error: Unable to determine type for file oS9kQIRuTq0Ya.auc.threshold.bed

The bed file in question either seems to be either corrupted or something in binary. image

mpmeers commented 2 years ago

Hello,

I'm guessing this is an input issue because there are no compression steps in the SEACR code--is it possible you're inputting a control bigwig instead of a bedgraph? It's possible that wouldn't be caught.

Mike