Using Majiq-CLIN pipeline to quantify splicing differences in Trio RNAseq

[Abhay Rastogi]

The Biociphers Team,

I'm a graduate student trying to run majiq V3. I found Dina Issakova's ISMB talk and Medrxiv paper (Aicher et al.). I aim to find differential splicing patterns between an affected proband and their parents using trio RNAseq data from the Undiagnosed Disease Network. Is the MAJIQ-CLIN pipeline now publicly available for academic use?

I'm trying to determine which commands to run for this analysis. I've followed the v3 command builder tool: https://biociphers.bitbucket.io/majiq-docs/commandbuilder.html

but I reckon the output commands are not accurate for my purposes. Is there a place I can find relevant documentation specific for clinical use case? 

I would also like to clarify if I am only interested in a specific gene, is it possible to get majiq to run only for that gene?

Best,

Abhay

[bsl...@seas.upenn.edu]

Dear Abhay,

MAJIQ-CLIN is not yet generally available, however:

(1) If you are non-commercial (e.g. part of a university lab) and want to use MAJIQ-CLIN now, please have your lab’s PI contact Yoseph Barash at yos...@upenn.edu with a description of the use-case and request for access. We will provide early access for academic use-cases intended to identify aberrant splicing between patients and controls.

(2) MAJIQ-CLIN will be generally available for academic and commercial licensing after the paper is published, which we hope will be soon.

If neither of the above works well for you, there are a couple more options:

(3a) Using MAJIQ V3 (available today!), run majiq psi and check (using your own analysis code) where the case (proband) E[PSI] is outside the range of the control (parents) E[PSI] with some minimum absolute PSI difference threshold (e.g. 0.2). Or, run majiq heterogen to compare a single “patient” (proband) against a group of controls (parents), but with groups this small you will need to carefully interpret the p-value output (today, MAJIQ V3 does not output raw scores for TNOM or other tests, only p-values). Neither of these get exactly the same result as MAJIQ-CLIN but it's a close approximation.

(3b) “Manual MAJIQ-CLIN”. MAJIQ-CLIN is a pipeline which takes user-provided configuration and then executes a series of MAJIQ V3 commands, so using only MAJIQ V3, it’s possible to do what MAJIQ-CLIN does. However, doing this manually is a bit involved and error-prone and we don’t really encourage or support this. If none of the above work for you, feel free to ask how to do this, with the caveat that we don’t encourage doing it this way and can provide only limited support.

Regarding your last question about how to run MAJIQ for just one gene: today there is no MAJIQ option to do this, but it’s possible by constraining the gff3 annotation to your one gene of interest. We recommend filtering the gff3 to your gene of interest using a tool such as GffRead and then executing MAJIQ with your new, smaller gff3 file. If you create your own gff3-filtering script rather than using an existing tool like GffRead, take care to retain all features within your gene by tracing the feature hierarchy up to the gene level (to ensure e.g. all exons are included). Another option is to instead filter your RNAseq bam files to reads contained in your gene of interest, for example with a command like $samtools view in.bam chr:start-stop . The MAJIQ execution is more computationally efficient using the former option (filter the gff3 rather than the bamfiles).

Best Regards,

Barry