Drug-response studies play an important role in both preclinical and clinical research, but such studies are complicated by differences in cell growth rates across samples and conditions. To improve the value and reliability of such studies, new metrics for parameterizing drug response were developed and published in Nature Methods by Marc Hafner, Mario Niepel, and Peter Sorger of the Harvard Medical School (HMS) LINCS Center. These new metrics, such as GR50 and GRmax, are derived from normalized growth rate inhibition (GR) values which are based on the ratio of growth rates in the presence and absence of perturbagen. Largely independent of cell division rate and assay duration, GR metrics are more robust than IC50 and Emax for assessing cellular response to drugs, RNAi, and other perturbations in which control cells divide over the course of the assay.

Future Updates and Improvements

We plan to continue adding features and improvements to the GRbrowser, GRcalculator, and the GRmetrics R package in the coming months. We welcome comments and suggestions at You can find a preliminary outline of our plans here. We will be adding a more detailed roadmap of additions/improvements in the near future.

Learn More about GR metrics Upload and analyze your own data Browse datasets analyzed using GR metrics

For offline computation, analysis, and visualization, see the Bioconductor R package GRmetrics.


Hafner M*, Niepel M*, Chung M, Sorger PK. (2016) Growth rate inhibition metrics correct for confounders in measuring sensitivity to cancer drugs. Nat Methods, vol. 13, 521–7. doi: 10.1038/nmeth.3853.

Hafner M, Niepel M, Sorger PK. (2017) Alternative drug sensitivity metrics improve preclinical cancer pharmacogenomics. Nat Biotech, vol. 35(6), 500-2. doi: 10.1038/nbt.3882.

Clark N*, Hafner M*, Kouril M, Williams EH, Muhlich J, Pilarczyk M, Niepel M, Sorger PK, Medvedovic M. (2017) GRcalculator: an online tool for calculating and mining dose-response data. BMC Cancer, in press.

Website design and online tool development

Nick Clark1, Marc Hafner2, Michal Kouril1, Mario Niepel2, Elizabeth Williams2, Jeremy Muhlich2 and Mario Medvedovic1

1 LINCS-BD2K Data Coordination and Integration Center; 2 HMS LINCS Center, Harvard Medical School