Giter Site home page Giter Site logo

Benchmarking SUPPA & rMATS about suppa HOT 11 CLOSED

comprna avatar comprna commented on August 13, 2024
Benchmarking SUPPA & rMATS

from suppa.

Comments (11)

EduEyras avatar EduEyras commented on August 13, 2024

from suppa.

JamalELH avatar JamalELH commented on August 13, 2024

Dear Eduardo,
Thank you so much for your quick and prompt reply!

What about A5' and A'3SS events that are also identically defined by rMATS and SUPPA?
Another related question regarding this benchmark, is it still "correct" to do so as rMATS is run directly on the alignment bam files "from STAR/Tophat" while SUPPA2 takes as inputs the expression at isoform level obtain from those bam files e.i they are not run on the same exact input??

Just for your information, I also run SUPPA2 after using Salmon vs RSEM vs Kallisto for the quantification at isoform level. I ve done this to see whether SUPPA2 depends on the quantification tool used. From the results I can say yes, which is normal as each of the three quantification tools has its own "method" to attribute reads to their corresponding isoforms, which may consequently influence the dispersion between replicates.

Bests regards,
Jamal.

from suppa.

EduEyras avatar EduEyras commented on August 13, 2024

from suppa.

JamalELH avatar JamalELH commented on August 13, 2024

Thank you so much for your comments!
I will keep you updated about the comparison of Salmon with Kallisto and rMATS.
In case I would like to share some results with you, where can I send it in private?

Best regards,
Jamal.

from suppa.

EduEyras avatar EduEyras commented on August 13, 2024

from suppa.

JamalELH avatar JamalELH commented on August 13, 2024

Dear Eduardo,
As a follow-up on our last discussion 'comparing SUPPA vs rMaATS outputs', the scenarios I found are the following:
SE events : rMATS detected 89999 while SUPPA detected 39090 event. Overlapping them, regardless of whether they were significant or not I found 52110 event (24.1%) detected by both, 13980 (13.4%) exclusively detected by SUPPA and 64889 (62.4%).
Significant SE events: 526 SE events were significant according to rMATS, 174 SE events according to SUPPA, only 20 (3%) were significant according to both.

I also overlapped A3/5SS and the results are quit similar to above.
Is this results seems OK according to your expertise with the above-mentioned pipelines?
Thank you so much in advance!
Please note that I used the same genome annotation files both SUPPA and rMATS pipeline (gencode.v27).
command lines:
SUPPA:
python $SUPPA diffSplice -m empirical -gc -s -me -th 1 -i $SUPPA_outputs/all_local_events.ioe -p $SUPPA_outputs/siRNA/siRNA_ERa.psi $SUPPA_outputs/siRNA/siRNA_CTRL.psi -e $SUPPA_outputs/siRNA/siRNA_ERa.tpm $SUPPA_outputs/siRNA/siRNA_CTRL.tpm -o $SUPPA_outputs/siRNA/siRNA_diffSplice.

from suppa.

EduEyras avatar EduEyras commented on August 13, 2024

from suppa.

JamalELH avatar JamalELH commented on August 13, 2024

Dear Eduardo,
Thank you for your quick reply!
My two case studies datasets are 80M and 30M respectively (each is 2 conditions in triplicates).
I wanted to be more exhaustive with rMATS so I did not use a cutoff but considered all those with a significant (FDR).

When I also used SUPPA but changed the quantifier upstream of it (salmon, rsem, kallisto), I found 100% overlap between the three when considering the events coordinates. But they differ in significance. Please refer to the attached venndiagram for more details about this point.

As I would like also to overlap with MAJIQ, which is able to detect more complex, non-binary local splicing events + de novo junctions/exons, is there a way to make a common data structure for the outputs from SUPPA and MAJIQ. Should I also consider the p-value column in SUPPA output as significance to perfom an experimental validation? because I knew that this p-value is for within genes events, not for between genes?

Thanks a lot for your help!
overlaps_je

from suppa.

EduEyras avatar EduEyras commented on August 13, 2024

from suppa.

ArashDepp avatar ArashDepp commented on August 13, 2024

Hi,

If you dont mind, could you please share the script for making SUPPA vs. Majiq comparison possible (as you mentioned in your comment). We used Kallitso + SUPPA2 for our research and need to defend this choice using Majiq for one of the reviewer.

from suppa.

EduEyras avatar EduEyras commented on August 13, 2024

from suppa.

Related Issues (20)

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. 📊📈🎉

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google ❤️ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.