CCLE drug response data

Question

[Mike]

Hello Biostars,

What is standard protocol to divide CCLE drug response data into two class; cell line sentivity and resistant, please also give reference of that protocol.

Thanks

Answer 1

[Ar]

Typically it is IC50 value. It is the concentration of an inhibitor where the response (or binding) is reduced by half. The IC50 is obtained by plotting a measure of the cell death rate vs the concentration of the drug. IC50 is expressed as a concentration so the higher the concentration of drug needed to kill the cell line the more resistant that cell line is. If the IC50 is very low then the concentration (or amount) of drug needed to kill the cell line is also very low which means the cell line is more sensitive. Typically you get a S curve or sigmoidal curve. IC50 is the middle point of the sigmoidal curve. The threshold for calling a drug sensitive/resistant changes depending upon the IC50 value and you need to see how the sigmoidal curve.

However, recent publications from John Quackenbush's lab has shown that IC50 values are not a good measure of calling a drug resistant or sensitive. I am not an expert but these are good starting point to understand the problems and possible measures:

1. http://f1000research.com/articles/5-825/v1
2. http://biorxiv.org/content/biorxiv/early/2015/09/06/026153.full.pdf
3. https://www.broadinstitute.org/ccle/home

[Mike]

Thanks Ar,

I also found other criteria based on activity area (ActArea, collumn in CCLE).

reference http://www.ncbi.nlm.nih.gov/pmc/articles/PMC4485860/

" For each drug, cell lines with normalized activity area at least 0.8 standard deviations (SDs) above the mean were defined as sensitive to the compound, whereas those with activity area at least 0.8 SDs below the mean were defined as resistant. Cell lines with activity area within 0.8 SDs of the mean were considered to be intermediate and eliminated from our analysis"

So, that Im not sure which one is standard criteria.