MAJIQ-L (long read data) help

[Tayvia Brownmiller]

Hello all,

Has anyone used the new long-read application? I am trying to implement it using IsoSeq outputs from my pacbio and I am struggling with what is needed regarding the long read tsv input.

Thank you all and thanks to the developers for the long read applications update!

[Tayvia Brownmiller]

Just wanted to provide an update. I was able to run the IsoQuant pipeline as mentioned in the MAJIQ-L preprint, but I am still unclear on what was specifically used as the input to MAJIQ-L as isoquant generates multiple tsv files. So far, with the files I have attempted to use I have not been successful at adding the long read splicegraph to voila view along side my short read data.

Any help in the matter would be greatly appreciated as I am currently structuring a manuscript with this data.

Thanks in advance!

[San Jewell]

Hi Tayvia,

Unfortunately, a message from a co-author of the MAjiq-L method tried to respond a few times but the google groups seem to have continued to lose his messages. Until we can get this resolved I am posting the reply from Seong below:

Thanks for your interest in MAJIQ-L! I apologize for the late response. To use the unified visualization in MAJIQ-L, you need two components, MAJIQ's outputs and LR tool's (IsoQuant in your case) outputs.

MQJIQ outputs: .psi voila and splicegraph

IsoQuant outputs: GTF file with discovered expressed transcript and TSV file indicating which reads contributed to transcript models

The former file ends with .transcript_models.gtf and the latter ends with .transcript_model_reads.tsv. Here is an example command option to generate .lr.voila file and visualize it.

voila lr 

--lr-gtf-file /PAHT/TO/SAMPLE_ID.transcript_models.gtf 

--lr-tsv-file /PAHT/TO/SAMPLE_ID.transcript_model.reads.tsv 

--voila-file /PAHT/TO/SAMPLE_ID.psi.voila 

-sg /PATH/TO/splicegraph.sql 

-o OUTPUT_FOLDER/lr.voila

voila view 

/PATH/TO/splicegraph.sql 

/PAHT/TO/SAMPLE_ID.psi.voila 

/PATH/TO/SAMPLE_ID.lr.voila 

--host 0.0.0.0

I hope this helps and let me know if you have any other questions.

Best,

Seong

Let us know if it helps,

Thanks,

-San

[Tayvia Brownmiller]

Hi San,

Thank you!!

I received a private message from Seong with all the help I was looking for. 

Perhaps the messaging issue is due to "reply to author" vs "reply all"? From my past experience working with groups like this that is usually the case.

Thanks again for your help and to the Biociphers team!

Tayvia

[San Jewell]

Glad to be of help and I hope the software can continue to be useful as we refine and update it.

Also for future readers of this thread, there is a minor error in the above message from Seong:

The former file ends with .transcript_models.gtf and the latter ends with .transcript_model_reads.tsv.

it should be

The former file ends with .transcript_models.gtf and the latter ends with .transcript_model_counts.tsv.

Thanks!

-San