Hi all, Encountering an issue today when trying t...
# crisprseq
i
Hi all, Encountering an issue today when trying to run `MAGECK_GRAPHRRA`:
Copy code
Cannot connect to ExperimentHub server, using 'localHub=TRUE' instead
Using 'localHub=TRUE'
If offline, please also see BiocManager vignette section on offline use
Error: package or namespace load failed for 'MAGeCKFlute':
.onLoad failed in loadNamespace() for 'depmap', details:
call: h(simpleError(msg, call))
error: error in evaluating the argument 'x' in selecting a method for function 'query': Invalid Cache: sqlite file
Hub has not been added to cache
Run again with 'localHub=FALSE'
Execution halted
Task failed
Has the
MAGECK_GRAPHRRA
process been validated to successfully work? I've attached my design file and contrasts to help inform further. I am attempting to run this on AWS HealthOmics with the latest version of the pipeline (2.2.0). Happy to provide any other details that may be helpful. I also note the pattern required in the design file to get RRA to work with replicates is a bit odd (as you can see, my "sample" and "condition" columns are identical, otherwise I believe RRA will fail given my design--but if there is another way to set this up that should work, please do let me know). Thanks!!
👀 1
Relatedly, the way the
rra
parameter is currently worded in the documentation makes it seem like either MAGeCK MLE or MAGeCK RRA can be run. Is this the case? I'm under the impression it should be possible to run both if one desires. Thanks!!
l
hi @Ittai Eres i unfortunately haven’t had time to have a look at this issue for now but hope to get back to you soon
@Ittai Eres could oyu please give me your command line?
i
No worries! Appreciate your efforts here. This was run from AWS HealthOmics, so I did not directly use the command line, but you can see the parameters I used in the attached JSON, and see the full engine log attached as well. Thanks!
Hi @Laurence Kuhlburger, any update here? Any other info or clarification I could provide? Or any example design you have where both RRA and MLE can be run with the pipeline, starting from FASTQ, and having replicates? Thanks!!
l
Hi Ittai, we added tests in a PR here https://github.com/nf-core/crisprseq/pull/185 because mageckflute is unfortunately causing a lot of issues. Could you however try running it with the dev version to see if our patch release worked?
Hi @Ittai Eres can you hit me up if you had time to try on the dev version? 🙂
i
Testing it now, will let you know what I see!
Hi @Laurence Kuhlburger, I implemented the dev version on these data, but still encountered the same error I pasted at the beginning of this thread. This time, on the MAGECK_FLUTEMLE_CONTRASTS process. Let me know if I can provide any more info or you have other tips on how to resolve, thanks!
l
Hi @Ittai Eres could you please send me the run + the log errors?
i
Hi @Laurence Kuhlburger, I've attached the logs I have access to--please let me know if I can provide more info. Thank you!!
l
Hi Ittai, i’m wondering whether i should make the FluteMLE module optional until i can figure it out. Is it generally a module that would be interesting to you? In the next PR you will also be able to run both rra and mle as you had once mentioned 🙂
i
Hi Laurence, thanks for circling back and for all your work on this. Making it optional for now for the sake of being able to run the pipeline seems reasonable, but I would note that ultimately this (MLE) is a functionality that I am interested in for this pipeline. Happy to do any testing with any new PR you recommend as well 😃
l
Hi Ittai, you’re also welcome to try to help out as maintaining this pipeline alone is starting to be quite challenging, we have a PR opened here : https://github.com/nf-core/crisprseq/pull/185