Majiq build finds 0 LSVs

[Yulia Ryvkin]

Hello!

I have an issue with majiq build:

I'm running 'majiq build' for an experiment with 2 conditions and 3 replicates each. Bam files are sorted and indexed. Genome and GFF3 - mm10.

Bam files were split from larger bam files of a single cell experiment. 

* The bam files are intact and functional, I've used them in IGV and also performed downstream analysis.

I got an ' Index 0' error while trying to run psi and dpsi.

In the process of identifying the cause, I found that the build step could not identify any LSVs.

Here is part of the build log:

2024-09-11 15:39:51,266 (PID:1861373) - INFO - Detect Intron retention GFP.DC

2024-09-11 15:40:37,799 (PID:1861373) - INFO - Done Reading file GFP.DC

2024-09-11 15:40:38,479 (PID:1861373) - INFO - Detecting LSVs ngenes: 41128

2024-09-11 15:40:48,812 (PID:1861373) - INFO - 0 LSV found

2024-09-11 15:40:48,813 (PID:1861373) - INFO - DUMP file b'LN_APCs.eTAC'

2024-09-11 15:40:49,484 (PID:1861373) - INFO - Create majiq file

2024-09-11 15:40:49,504 (PID:1861373) - INFO - Dump majiq file

2024-09-11 15:40:49,514 (PID:1861373) - INFO - LN_APCs.eTAC: 0 LSVs

2024-09-11 15:40:49,514 (PID:1861373) - INFO - DUMP file b'ZsGreen.eTAC'

2024-09-11 15:40:50,900 (PID:1861373) - INFO - Create majiq file

2024-09-11 15:40:50,937 (PID:1861373) - INFO - Dump majiq file

2024-09-11 15:40:50,942 (PID:1861373) - INFO - ZsGreen.eTAC: 0 LSVs

2024-09-11 15:40:50,942 (PID:1861373) - INFO - DUMP file b'GFP.eTAC'

2024-09-11 15:40:53,694 (PID:1861373) - INFO - Create majiq file

2024-09-11 15:40:53,772 (PID:1861373) - INFO - Dump majiq file

2024-09-11 15:40:53,777 (PID:1861373) - INFO - GFP.eTAC: 0 LSVs

It seems highly unlikely that there are no differences, since these are two separate cell populations. 

Could you please help me resolve this issue?

here is my code:

majiq_2.5.7.sif majiq --license /apps/containers/singularity/Majiq/majiq_license_academic_official.lic \

build $Majiq_output/ensembl.mm10.gff3 \

--conf $Majiq_output/configuration_mm10.txt \

--output $Majiq_output/zsgreen_build_ensembl

and my configuration file content:

[info]
bamdirs=/path/to/Majiq/Bams
genome=mm10
genome_path=/path/to/mm10/Sequence/WholeGenomeFasta
[experiments]
eTAC=LN_APCs.eTAC.sorted,ZsGreen.eTAC.sorted,GFP.eTAC.sorted
DC=LN_APCs.DC.sorted,ZsGreen.DC.sorted,GFP.DC.sorted

Best regards,

Yulia

[San Jewell]

Hi Yulia,

There are two common causes I've seen that can lead to this, so I would check them first:

1) Check that the chromosome names in the gff3 file and the BAM file are matching

2) Check that the strandness of the experiments is correctly reflected in the build config file

Let me know if either of these investigations lead you to have nonzero LSVs found.

Thanks,

-San

[Julia R]

Hi San,

You were correct, it was the chromosome naming issue.

Thank you so much for your help!

Best,

Yulia

--
You received this message because you are subscribed to the Google Groups "Biociphers" group.
To unsubscribe from this group and stop receiving emails from it, send an email to majiq_voila...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/majiq_voila/ccaf2091-0dad-4be2-9fd1-aa79744fda67n%40googlegroups.com.

--

Julia Ryvkin, Ph.D. 

+972-54-6148944

You may get emails from me outside of normal working hours. Please do not feel any pressure to respond outside of your own working pattern.

[San Jewell]

Hi Yulia,

I'm glad it was easy to solve!

Best,

-San