Another issue, and I don't get this one at all - pipe generates the following matrix (which looks fine to me) "Samples baseline hypoxia1_hypoxia2_vs_control1_control2 control1 1 0 control2 1 0 hypoxia1 1 1 hypoxia2 1 1" but is causes this error, works fine with only one hypoxia sample, (i.e. control1 1 0 control2 1 0 hypoxia1 1 1) but for obvious reasons I want to use both Error below for using 2 samples

Command error:
 INFO @ Wed, 09 Oct 2024 18:29:54: Parameters: /usr/local/bin/mageck mle --threads 6 -k count_table.count.txt -n hypoxia1,hypoxia2_vs_control1,control2 -d hypoxia1_hypoxia2_vs_control1_control2.txt 
 INFO @ Wed, 09 Oct 2024 18:29:55: Cannot parse design matrix as a string; try to parse it as a file name ... 
 INFO @ Wed, 09 Oct 2024 18:29:55: Design matrix: 
 INFO @ Wed, 09 Oct 2024 18:29:55: [[1. 0.] 
 INFO @ Wed, 09 Oct 2024 18:29:55: [1. 0.] 
 INFO @ Wed, 09 Oct 2024 18:29:55: [1. 1.] 
 INFO @ Wed, 09 Oct 2024 18:29:55: [1. 1.]] 
 INFO @ Wed, 09 Oct 2024 18:29:55: Beta labels:baseline,hypoxia1_hypoxia2_vs_control1_control2 
 INFO @ Wed, 09 Oct 2024 18:29:55: Included samples:control1,control2,hypoxia1,hypoxia2 
 INFO @ Wed, 09 Oct 2024 18:29:55: Loaded samples:control1;control2;hypoxia1;hypoxia2 
 INFO @ Wed, 09 Oct 2024 18:29:55: Sample index: 2;3;0;1 
 INFO @ Wed, 09 Oct 2024 18:29:55: Loaded 182 genes. 
 Traceback (most recent call last):
  File "/usr/local/bin/mageck", line 66, in <module>
 Error loading line 218
 Error loading line 521
 Error loading line 907
   main();
  File "/usr/local/bin/mageck", line 43, in main
   args=crisprseq_parseargs();
  File "/usr/local/lib/python3.9/site-packages/mageck/argsParser.py", line 258, in crisprseq_parseargs
   mageckmle_main(parsedargs=args); # ignoring the script path, and the sub command
  File "/usr/local/lib/python3.9/site-packages/mageck/mlemageck.py", line 74, in mageckmle_main
   allgenedict=read_gene_from_file(args.count_table,includesamples=args.include_samples)
  File "/usr/local/lib/python3.9/site-packages/mageck/mleinstanceio.py", line 84, in read_gene_from_file
   ginst.nb_count=np.matrix(ginst.nb_count)
  File "/usr/local/lib/python3.9/site-packages/numpy/matrixlib/defmatrix.py", line 145, in __new__
   arr = N.array(data, dtype=dtype, copy=copy)
 ValueError: setting an array element with a sequence. The requested array has an inhomogeneous shape after 1 dimensions. The detected shape was (4,) + inhomogeneous part.

Could you show me your samplesheet too?

Sure

could you try without the underscores in condition? I wonder if that’s the issue

Yup and I’ve tried without too. It’s more to do with design matrix size. 1v2 works but 2v2 and bigger fails.

the error is a mageck mle one and not really proper to crisprseq somehow

Google suggests it might be a numpy version issue? But that’s in the docker right, so not something I am influencing

so just using mageck RRA?

RRA works fine. MLE has this error

Okay, if you have time would you mind opening an issue on the github repo? I’m a bit swamped so can’t look too much into it but it’d be good to keep track of it

Okay. Happy to do that.

did you by any chance try running mageck outside of the pipeline and see if you get the same error?

I tried, but I got a whole set of different issues that were to do with the python and the local environment. Which fixing wasn’t going to solve this problem. Ultimately the fix there is to get mageck install by the team the runs the cluster as installing the into our home directory is less than ideal. And of course what nextflow should avoid

Yes makes sense. I was just wondering whether the issue is related to the mageck version, because this is more a software error than a crisprseq one but i agree with you that the design you made should run through so i’m a bit unsure on how to help

My guess is that numpy in the docker is the version which is known to have a bug that gives this error. But I that was a rabbit hold I haven’t gone down. As I assume the docker was tested.

yes and the docker is made with the mageck package which normally should have the right versions of numpy

Does the mageck package have it’s own issues page? I could check that.

Thanks! I'll try to reproduce it but if it's a mageck issue I think there's not much else I can do

I couldn't find a specific place to put an issue on the Mageck MLE module.

Ah did you add it on the mageck bit bucket or on the crisprseq github?

crisprseq is the issue I added today.

Hi Following up on this, I have the same issue. Wondering if you were able to fix it? thanks!