Discrepancies between downloaded files and GUI

Q&A
1

I see a large difference between the number of distinct(ModelID) in the CRISPRGeneEffect.csv versus the Model.csv download from here. This misalignment doesn’t appear using the GUI Data Explorer.
Is there a reason for this mismatch? How can we download the complete list of ModelID with CRISPR gene effect?
Thanks

2

Hi,

These two files don’t have the same number of ModelIDs because not all models have CRISPR screens. Some models we release only have omics data, and the metadata for these are also included in Model.csv.

Thanks,
Simone

1 Like
3

The Model.csv file contains model IDs for all models that we have represented in the DepMap portal. This includes data types which were generated elsewhere and includes data types other than the CRISPR screens. (For example, the Sanger drug sensitivity screening data)

If you want the set of ModelIDs which have CRISPR gene effect values, I’d simply take the list from the row headers of CRISPRGeneEffect.csv. That is the definitive source of gene effect for all models for which we have CRISPR screening data. (Where some of that data was generated within the DepMap project, and some has come from other sources and computationally integrated into that dataset)

Thanks,
Phil

1 Like
4

Thank you for your responses. I checked the DepMap GUI and cell lines appear on the “Data Explorer” that are not in the download. Is there a different download that would include all the CRISPR KO Scores that are available on the GUI?

5

Can you attach a screenshot or provide an example of a cell line which you say is present in Data Explorer, but missing from the CRISPRGeneEffect.csv file?

The CRISPR data you see in Data explorer and CRISPRGeneEffect.csv should be consistent, so an example of where they’re not would be very helpful to track down what you’re seeing.

Thanks,
Phil

1 Like