I received a question via email regarding the values in the dose-response files:
>Could you please clarify for me the meaning of the values?
The first column indicates the concentration of the test compound used for each row. This is in the format: -log10([M]) where M is the molar concentration of the test compound. Larger values indicate "more drug", but on a log 10 scale.
The remaining values are the percent viability of the cells (when using the compound - indicated by the column header - at the concentration indicated for that row).
Here's an example protocol for how these measurements can be obtained and calculated experimentally (note that this data was not generated with an MTT assay and that there are lots of cell viability assays that could be used, but the general principle is the same): https://blog.quartzy.com/2017/05/01/cell-viability-assays-mtt-protocol
Created by Robert Allaway allawayr And here is a follow-up set of questions on this same topic:
>1. If the values are the percent of cells viable after applying the given drug concentration, what do values above 100% mean? Does it mean that the cells actually "multiplied" after treatment with the drug? Would such a drug be desirable as a treatment option?
Values above 100% can be the result of the following: experimental noise, a true increase in cell division rate (not common in my experience), increase in metabolic rate (causing increased metabolism of the assay reagent for assays dependent on that), assay interference from the test compound (for example - some nanoparticles are [documented](https://pubmed.ncbi.nlm.nih.gov/21090919-gold-nanoparticle-trafficking-of-typically-excluded-compounds-across-the-cell-membrane-in-jb6-cl-41-5a-cells-causes-assay-interference/) as being able to interfere with cellular toxicity assays).
>2. Is the I20 dose for each drug among the tested concentrations in the first column? If yes, would you be able to tell us which one is it for each drug and each cell line?
No, the IC20 is calculated by fitting a curve to the full set of dose-response values for a given compound-cell line pair. GraphPad has a nice writeup of how this is (more generally) calculated: https://www.graphpad.com/support/faq/50-of-what-how-exactly-are-ic50-and-ec50-defined/
If you want to calculate the IC20 yourself, you can use a package such as R's `nplr` to calculate it. There are python packages like `ECCpy` that can probably do it too, but I haven't used them myself!
>3. In the webinar, you say the I20 dose is at 48 hours but the gene expressions corresponding to it were measured at 24 hours, does that mean you've used an outside provided I20 value rather than determining it yourself?
See above- this was calculated by fitting a curve to the full set of dose-response measurements for a given compound-cell line pair after 48 hours of compound treatment. This IC20 concentration was then applied to the cell line for 24 hours, after which the cells were PLATE-seqed.
I presume @efd2115 used a four-parameter logistic regression or similar to calculate the IC20 - tagging him here in case this isn't correct.
Cheers,
Robert