Run docker pipeline4.0 but receive 0 events

[Chris-ZhangZ]

Hello developers: I came across some troubles when running the following codes with psi_sigma_pipeline4.0 from docker.

~$ cd /Volumes/Untitled/2312_STAR_out_DC1/work

~$ docker run --mount src=pwd,target=/Volumes/Untitled/2312_STAR_out_DC1/work,type=bind -it docker.io/woodydon/psi_sigma_pipeline:4.0

~$ cd /Volumes/Untitled/2312_STAR_out_DC1/work

~$ ls

1_Control.SJ.out.tab 2_Mutant.SJ.out.tab 4_Control.SJ.out.tab 1_Control.bam 2_Mutant.bam 4_Control.bam 1_Control.bam.bai 2_Mutant.bam.bai 4_Control.bam.bai 1_Mutant.SJ.out.tab 3_Control.SJ.out.tab 4_Mutant.SJ.out.tab 1_Mutant.bam 3_Control.bam 4_Mutant.bam 1_Mutant.bam.bai 3_Control.bam.bai 4_Mutant.bam.bai 2_Control.SJ.out.tab 3_Mutant.SJ.out.tab Mus_musculus.GRCm38.100.sorted.gtf 2_Control.bam 3_Mutant.bam groupa.txt 2_Control.bam.bai 3_Mutant.bam.bai groupb.txt

(The input files include .bam .bam.bai and *.SJ.out.tab generated by STAR, together with groupa.txt and groupb.txt containing .bam file fullnames as required.)

~$ perl /usr/local/bin/PSI-Sigma-2.3/dummyai.pl --gtf Mus_musculus.GRCm38.100.sorted.gtf --name YTHDC1_KO --nread 10 --type 1 --fmode 3 --irmode 2 --threads 2 --groupa groupa.txt --groupb groupb.txt

Recorded:

... Checking SJ reads... Calculating PSI values... Number of events = 0 Number of samples = 8 Statistics option = Student's t-test number of p-value = 0 Number of final p-value = 0 Skipping p-value adjustment. number of fdr(BH) = 0 ===PSI analysis spent 0.0114 hours.=== Filtering ΔPSI results... Filtering mode = 3 Reading... Mus_musculus.GRCm38.100.sorted.gtf.mapping.txt Reading... YTHDC1_KO.db ===Filtering spent 0.0003 hours.=== Archiving... YTHDC1_KO_r10_ir3.txt YTHDC1_KO_r10_ir3.txt

Troubles: The output number of events is 0 as reported, although PSI-sigma2.3 seems to be running correctly. I tried several times with some modifications through renaming files, regenerating .bam/.bam.bai/.SJ.out.tab files or resorting .gtf but with no optimization. Meanwhile, I found the samtools version is 0.1.19 (in the docker container), which does not agree with the information in the workshop (Day1.ppt, Noting that samtools==1.11 is recommended for PSIsigma env build). Could it be the version issue of samtools that cause the problem?

Looking forward to your reply !

Chris Zhang

[wososa]

Hi @Chris-ZhangZ ,

Thanks for trying PSI-Sigma. Could you please check if groupa.txt and groupb.txt is like this format?

1_Control 2_Control 3_Control 4_Control

Also, another common cause is the Chromosome names. could you please check if 1_Control.SJ.out.tab is using exactly the same Chromosome names as the Mus_musculus.GRCm38.100.sorted.gtf? For example, are the two files using "1" for Chromosome 1? chr1 will not be considered as equal to 1.

Best, Woody

[Chris-ZhangZ]

Thank you @wososa for your advice, The problem has been solved now after changing the chromosome names in Mus_musculus.GRCm38.100.sorted.gtf ("" to "chr" makes PSI-sigma work out).

Warm Regards Chris Zhang