About negative numbers in the "spliced_with_coord" column in the Voila Modulizer output

[Shengwei Xiong]

Hello all,

I encountered an issue while working with the Voila Modulizer output, specifically related to negative numbers in the "spliced_with_coord" column. Are those simply bad lines?

Best,
Shengwei

[San Jewell]

Hi Shengwei,

I might need to see the rest of the file but I'm guessing the exon in these lines is a half exon, and thus doesn't have a start coordinate. If this is not the case then it sounds like it might be a bug. Can you confirm? See here for information about half exons: https://biociphers.bitbucket.io/majiq-docs-academic/getting-started-guide/faq.html

Thanks,

-San

[Shengwei Xiong]

Hi San,

I don't think it is a half-exon problem. I have attached the TSV file. Could you take a look? Thank you!

Best, 

Shengwei

[San Jewell]

Hi Shengwei,

I recently ran modulizer over one of our larger sets of gtex cancer data as a wide-scope test, using the same arguments you provided. Looking at alternate first exon output I wasn't able to reproduce this kind of issue / result. Therefore I feel that there must be some kind of weird unexpected splicegraph situation happening in the data that you are working with. Based on your security / privacy principals, would you have any issue sharing your data set (splicegraph.sql, sj/majiq/voila files) with me for debugging purposes? I can provide a secure place to sftp upload the data if you decide to proceed.

Thanks,

-San