Standard for representing CRISPR data, Chronos

Dear DepMap experts, I am using DepMap to show the dependency of various cancer to LDLR. I have following questions:

  1. What is the standard of reporting Chronos number, I mean is there a level of significance we can report? For example, is 0.3 Gene effect (Chronos) number quite good to report as a cancer dependence gene or numbers beyond -1 and lesser are more powerful and should be reported? I do see a red line in the graph at -1 position only, is that a minimum threshold a gene should show to pass as a cancer dependency gene or any number below 0 is also good?
  2. Also, to qualify as a pan essential gene, the Chronos number should be -1 or below always? For example in the attached graph LDLR could qualify as a pan essential gene in the blood cancer but not in the others??
  3. For example in Cervix cancer, the expression of LDLR is high (TPM~5) but the gene effects number is low (only 0.2), does it mean although the level of LDLR is high in cervix cancer but its knock out does not affect cancer cell viability. Not sure how to say it in terms of statistics, can we say it has significant affect on viability but moderate.
  4. Can I get a reference where DepMap was used to to show the importance of a single gene, like I want to show the importance of LDLR. This would help me to write my results.
    Thank you for making such a nice database. I am looking forward to hear back from you.
    This is link for LDLR: LDLR DepMap Gene Summary


Hello VJ,

There are some tooltips in the portal that can help in answering your questions, for example how we decide if a cell is dependent:

And the meaning of -1 gene effect:

Screen Shot 2022-01-25 at 5.02.15 PM

Dependency probabilities can be downloaded from the downloads page (file names ending in “gene_dependency”). But it is up to you to define what you want to consider a dependency for your research.

Pan-essential by definition means essential for all lines (although we allow for some lines not to show dependency to allow for random noise). How we determine which genes are pan-essential is based on how depleted they are in their 90th percentile least depleted cell line. LDLR would not qualify as a pan essential even within blood cancers.

Searching for references for you is outside the scope of support we provide. Each data release has its own DOI. You could try searching for papers that cite those, or looking through the many DepMap publications listed on the portal itself.

1 Like