--
You received this message because you are subscribed to the Google Groups "gsea-help" group.
To unsubscribe from this group and stop receiving emails from it, send an email to gsea-help+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/gsea-help/57e2dccf-5f4f-4e19-97ab-f1bd91f9e4e0n%40googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/gsea-help/CAGCeyZzk5_q-ir1rn3RQugQnCOnheCgVPryeZP_Dx1NXw%3DYWCA%40mail.gmail.com.
Hi Mahtab,
This result indicates that there was something wrong with the null distribution that was generated for this data.
This can occur if there is a large skew in the dataset causing there to be no valid null generated for the given set on the same side as the true score.
How many genes are in the ranked list, and on the enrichment plot where is the “Zero cross” annotated?
You can also look at the null ES distribution plot at the bottom of the set report page.
-Anthony
Anthony S. Castanza, PhD
Curator, Molecular Signatures Database
Mesirov Lab, Department of Medicine
University of California, San Diego
To view this discussion on the web visit https://groups.google.com/d/msgid/gsea-help/CAKqtSN3X3eoc-EPn3mJZR%2BiKt%3DCGT_La9Fx%2BUxA9AzUaCi-z%3DQ%40mail.gmail.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/gsea-help/SJ0PR05MB7609E4CF3169E01F062BC41BF7819%40SJ0PR05MB7609.namprd05.prod.outlook.com.
The zero cross should be annotated on all of the enrichment plots. Or it’s also the number of genes that have positive ranking metrics vs. the number of genes that have negative ranking metrics. A large skew in one direction or another can lead to this sort of issue.
Since you’re using preranked data you might also want to check the list for any genes with “Inf” “-Inf” “NaN”, etc values, these can be what the “Scoring produced infinite or NaNs values” warning can be referring to.
If you like, you can send your ranked list and the gene sets you’re analyzing to gsea...@broadinstitute.org and I can see if I can figure out what precisely is causing this to occur and if there is an easy fix for it.
To view this discussion on the web visit https://groups.google.com/d/msgid/gsea-help/CAKqtSN2aY_coisvsrE9tYjv6BUKEx35gcFZS4fLhGkOoYaW3Sg%40mail.gmail.com.
Hi Esther,
We generally ask that people create new topics for their specific issue to prevent undesired notifications to the people that originally created a given topic.
That said, it appears that GSEA is having difficulty parsing your TXT formatted expression dataset. If you could, please double check and confirm that your file matches the specification here for .txt formatted files: https://software.broadinstitute.org/cancer/software/gsea/wiki/index.php/Data_formats#TXT:_Text_file_format_for_expression_dataset_.28.2A.txt.29
Assuming that your file looks like it matches that format, and isn’t accidentally a .GCT file that has the wrong file extension, could you perhaps send a screenshot of the file open in a plain text editor? That would help us to diagnose what could be going wrong here.
One quick thing you might also try, assuming everything else looks good (although this shouldn’t be actually causing an error) would be to replace the text “Description” with “DESCRIPTION” (all uppercase, no quotes). The parser shouldn’t have trouble with the mixed-case version, but we’ve been working on the parser code recently so it’s always possible we’ve introduced some unintended behavior.
Thanks!
-Anthony
Anthony S. Castanza, PhD
Department of Medicine
University of California, San Diego
To view this discussion on the web visit https://groups.google.com/d/msgid/gsea-help/2b92765b-fed1-4c2b-be42-82e51ed28c1an%40googlegroups.com.