PSI doesn't run- the parameter window is presented insted

[eladzis]

Hello @wososa I'm running the same command on two different dirs (I just change the path of the files and run the common on different dirs). At the some directories it work and run normally and at the others it doesn't and I get the parameters data instead: INFO: Using cached SIF image

Usage: perl dummyai-parallel.pl [parameters] Parameters: --gtf [text] the gene annotation file for building PSI-Sigma database. --name [text] the prefix of PSI-Sigma database and output files. --type [number] 1: short-read RNA-seq data 2: long-read RNA-seq data --nread [number] the minimum number of supporting reads for a splicing event. --output [/path] Publish results to the designated folder. (default: current folder) --skipratio [number] the ratio (0~1) of skipping reads in a exon-skipping event [default: 0.05]. A splicing event has to have at least one input sample passing this criteria to be included in the analysis. --fmode [number] 0: delta-PSI > 10% and p-value < 0.01 (default/recommended) 1: delta-PSI > 10% 2: p-value < 0.05 3: report all events --irmode [number] 0: only introns in the .gtf. (default) 1: aggressively search for all introns 2: Skip intron-retention events. --irrange [number] 0: IR event is using reference points of the target exon region. n: Use +-n bases around target intron region and +-n bases around event region. (default) --irclean [number] 0: No action (default) 25~100: Apply skipratio and remove intron-retention events whose coverage difference between the its two splice sites is > n% of their average. --adjp [number] 0: Skip p-value adjustment. (default) 1: Benjamini-Hochberg (Statistics::Multtest Perl module) 2: Benjamini-Hochberg (qvalue or p.adjust() R package) --trimp [number] Set FDR to 'na' for events whose maximum PSI values are below [number]% or whose minimum PSI values are above [100-number]% in all groupa.txt and groupb.txt during p-value adjustment. [default:5] --denominator [number] 0: Don't report denominators. (default) 1: Report the table of denominators. --variance [number] 0: Assuming equal variance and use Student's t-test. (default) 1: Assuming unequal variance and use Welch's t-test. --groupa [/path/groupa.txt] Specify where the groupa.txt file is. (default: groupa.txt) --groupb [/path/groupb.txt] Specify where the groupb.txt file is. (default: groupb.txt) --threads [number] the number of CPUs used when --irmode > 0. 1: one CPU thread (4~8GB RAM per CPU thread) 1+: more CPU thread (2 = 8~16GB RAM, 3 = 12~24GB RAM, and so forth.)

I tried to run the tool on the same data but in different dirs. At one the parameter window didn't appear and at the other it did. Can you help me understand what might causes this?

Best,

Elad Zisman

[wososa]

Hi @eladzis ,

could you please paste your exact command? It is probably because of additional space or / sign.

Best, Woody

[eladzis]

I thought about that also. That is way I copied the command from where it did work. Also I tried running it manually, adding each parameter needed (first GTF, then when it told me you need to input nread I add that parameter, until no parameter was required but the parameter menu still appeared) This is an example of the command:

group_a="/sci/labs/milotem/eladzis/Michal_Loteme/SplicingAnalysis/groupa.txt"
group_b="/sci/labs/milotem/eladzis/Michal_Loteme/SplicingAnalysis/groupb.txt"

gtf_url="/sci/labs/milotem/eladzis/Michal_Loteme/SplicingAnalysis/psi_referance_geonme/GRCh37_genes.sorted.gft"
singularity exec --bind /sci:/sci docker://woodydon/psi_sigma_pipeline:3.9 perl /sci/labs/milotem/eladzis/Michal_Loteme/SplicingAnalysis/PSI-Sigma-2.1/dummyai.pl --gtf $gtf_url --groupa ${group_a} --groupb ${group_b} --nread 10 --name test_psi_sigma --type 1 --fmode 3 --threads 2

Best, Elad

[wososa]

Hi @eladzis ,

Thanks for sharing your command. Could you try replacing $gtf_url to /sci/labs/milotem/eladzis/Michal_Loteme/SplicingAnalysis/psi_referance_geonme/GRCh37_genes.sorted.gft? I wonder if the variables ($gtf_url,$group_a,$group_b) would make any difference. Or you can try add {} to $gtf_url.

This might be irrelevant. Is the file supposed to be named with ".gft" in the end? I wonder if the file name is correct.

Best, Woody

[eladzis]

Hi @wososa Thanks a lot, the end of the gtf file was the problem, Now it works perfectly

Best,

Elad