Incorrect number of cells in stroma tissue category

akoyabio / phenoptrReports

Create analysis reports from inForm data.

https://akoyabio.github.io/phenoptrReports

zlib License

15 stars 7 forks source link

Incorrect number of cells in stroma tissue category #15

Closed rvanguri closed 5 years ago

rvanguri commented 5 years ago

I am running the "Analyze consolidated data" app and I see that it does find any cells for a particular phenotype in the Stroma category. The Tumor tissue category calculates correctly though. I've tried variations where I try to calculate only in the Stroma, Tumor and Stroma and Tumor alone. The Tumor totals are always correct but Stroma always results in 0 cells. If I manually grep through the Consolidated_data.txt it is clear that there are cells in the Stroma tissue category that should get counted.

AB-Kent commented 5 years ago

Hi, can you possibly send me the Consolidated_data.txt and Script.R files that show this problem? Thanks!

rvanguri commented 5 years ago

Please see attached. phenoptr_issue.zip

AB-Kent commented 5 years ago

Thank you for the data! The problem is that a few of your cells do not have cytoplasm data. This caused the sum to be NA which was then mis-interpreted by the counting code.

A couple of solutions for you:

I've checked in a fix which you can get by installing the dev version with devtools::install_github('akoyabio/phenoptrReports@9000')
Modify your phenotype expression to check for NA values; the correct expression is CD3+/~!is.na(`Cytoplasm CD8 (Opal 690) Mean`) & `Cytoplasm CD8 (Opal 690) Mean`>0.65

Thanks again for the bug report and the data to repro!

rvanguri commented 5 years ago

This works perfectly now, thanks!