Hello! Running the pipeline using

mageck mle

I notice that the resulting

.sgrna_summary.txt

output file predicts all sgRNAs have an efficiency of exactly 1. This seems to indicate the sgRNA efficiency component of the pipeline is either failing or not being run at all. Is there any way to diagnose what is going on here? Does the pipeline actually have the ability to incorporate sgRNA efficiency predictions? My command is as below:

nextflow run nf-core/crisprseq \
    --analysis screening \
    --outdir results \
    --input samplesheet.csv \
    --library library.txt \
    -profile docker \
    --mle \
    --contrasts contrasts.csv

And the output of the

.sgrna_summary.txt

file looks like this:

Gene	sgRNA	eff
CREBBP	GCAAGGTCGGGCCAGACACC	1
CREBBP	GGCGCAGGCGCACTGCGATC	1

Hey this is more on the mageck side, i remember reading something about this in their documentaiton but don’t have a lot of time to look it up right now, maybe you could have a look https://sourceforge.net/p/mageck/wiki/advanced_tutorial/#:~:text=MAGeCK%20mle%20also%20allows%20an,the%20SSC%20program%20as%20indicated.

Hi @David Tatarakis, were you able to figure out what's. causing this issue where sgRNAs efficiency = 1? I am having the same problem. Thank you!

@Aastha Pal Haven't figured this out yet. I've read through the documentation that @Laurence Kuhlburger suggested before. In their tutorial, the MAGeCK folks have you run the

SSC

software separately and then provide the tabular file with gRNA names, sequences, and efficiencies as an input to the

mageck mle

command. However, it seems that this is run internally as a part of the nf-core pipeline since that file is provided as an output. The only thing I noticed was that

SSC

requires the first column of the

library.txt

file to contain the actual sgRNA sequence. So I made sure that was the case. However, no change to the output as a result. I suppose I could try running

SSC

myself separate from Nextflow. But I'm not sure if the nf-core pipeline is able to accept the sgRNAs efficiencies file as an input to

MAGeCK mle

. Would really love some guidance here!

Okay. its still the same. @Laurence Kuhlburger, Any suggestion?

Hey so SSC is a software external to MAGeCK? We could try adding it to nf-core/crisprseq, this isn't yet available