Feature Request: Inter-sample PSI Value Comparison for Large Datasets in SUPPA2

[XXuxi]

Dear SUPPA2 Development Team,

I hope this message finds you well. My team and I are avid users of SUPPA2 for splicing analysis, and we are currently working with larger sets of samples. Our goal is to compare PSI values produced by the psiPerEvent function across a substantial number of samples to calculate scores that reflect the variation likelihood for splicing events among these samples, ideally resulting in a comprehensive numerical matrix.

Is there any functionality within SUPPA2 that can assist in comparing the sample PSI values from psiPerEvent across multiple samples to derive a score, hence facilitating the creation of a large-sample numerical matrix?

If such an option is not currently available, could you advise on a potential approach or suggest any complementary tools that could accommodate our needs for inter-sample comparison of PSI values?

Your guidance in this matter would be invaluable to our research and any further understanding of variable splicing probabilities across different loci among numerous samples.

Thank you very much for your time and support.

Best regards,
Xi Xu

[EduEyras]

Dear Xi Xu, Thanks a lot for your email. If you have two groups, SUPPA provides a function to calculate the differential splicing between the two groups. It looks at each event PSI values across the two groups of samples and perform a statistical test. If you have multiple groups, you can combine all PSI values into one matrix and perform clustering with the cluster functions in SUPPA. If the groups are not known, you could select the events that have the largest variation across samples and use those to perform some unsupervised clustering. In the wiki pages there are some practical examples that might help: https://github.com/comprna/SUPPA/wiki/SUPPA2-tutorial I hope this helps Thanks for using SUPPA Eduardo

…

On Thu, 8 Feb 2024 at 03:18, XXuxi ***@***.***> wrote: Dear SUPPA2 Development Team, I hope this message finds you well. My team and I are avid users of SUPPA2 for splicing analysis, and we are currently working with larger sets of samples. Our goal is to compare PSI values produced by the psiPerEvent function across a substantial number of samples to calculate scores that reflect the variation likelihood for splicing events among these samples, ideally resulting in a comprehensive numerical matrix. Is there any functionality within SUPPA2 that can assist in comparing the sample PSI values from psiPerEvent across multiple samples to derive a score, hence facilitating the creation of a large-sample numerical matrix? If such an option is not currently available, could you advise on a potential approach or suggest any complementary tools that could accommodate our needs for inter-sample comparison of PSI values? Your guidance in this matter would be invaluable to our research and any further understanding of variable splicing probabilities across different loci among numerous samples. Thank you very much for your time and support. Best regards, Xi Xu — Reply to this email directly, view it on GitHub <#181>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ADCZKB4BB5MJYOPANUUMYYDYSOSLZAVCNFSM6AAAAABC6DYTROVHI2DSMVQWIX3LMV43ASLTON2WKOZSGEZDGMZZG4YDSNI> . You are receiving this because you are subscribed to this thread.Message ID: ***@***.***>