Skip to main content

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

Alternative drug sensitivity metrics improve preclinical cancer pharmacogenomics

Nature Biotechnology volume 35, pages 500–502 (2017)Cite this article

Subjects

To the Editor:

This is a preview of subscription content, access via your institution

Relevant articles

Open Access articles citing this article.

Access options

Rent or buy this article

Prices vary by article type

from$1.95

to$39.95

Learn more

Prices may be subject to local taxes which are calculated during checkout

Figure 1: Division rate confounds IC50 values and pharmacogenomic associations.
Figure 2: Variation in efficacy and genomic enrichment based on GRmax values in the gCSI data set.

References

  1. Hafner, M., Niepel, M., Chung, M. & Sorger, P.K. Nat. Methods 13, 521–527 (2016).

    Article  CAS  Google Scholar 

  2. Haibe-Kains, B. et al. Nature 504, 389–393 (2013).

    Article  CAS  Google Scholar 

  3. CCLE Consortium & GDSC Consortium. Nature 528, 84–87 (2015).

  4. Harris, L.A. et al. Nat. Methods 13, 497–500 (2016).

    Article  CAS  Google Scholar 

  5. Niepel, M., Hafner, M., Chung, M. & Sorger, P.K. Curr. Protoc. Chem. Biol. 9, 55–74 (2017).

    Article  CAS  Google Scholar 

  6. Hafner, M., Niepel, M., Subramanian, K. & Sorger, P.K. Curr. Protoc. Chem. Biol. 9, 96–116 (2017).

    Article  CAS  Google Scholar 

  7. Haverty, P.M. et al. Nature 533, 333–337 (2016).

    Article  CAS  Google Scholar 

  8. Im, H.K. et al. PLoS Genet. 8, e1002525 (2012).

    Article  CAS  Google Scholar 

  9. Choy, E. et al. PLoS Genet. 4, e1000287 (2008).

    Article  Google Scholar 

  10. Eichhorn, P.J. et al. Cancer Res. 68, 9221–9230 (2008).

    Article  CAS  Google Scholar 

  11. Rozenblatt-Rosen, O. et al. Mol. Cell. Biol. 25, 612–620 (2005).

    Article  CAS  Google Scholar 

  12. Fallahi-Sichani, M., Honarnejad, S., Heiser, L.M., Gray, J.W. & Sorger, P.K. Nat. Chem. Biol. 9, 708–714 (2013).

    Article  CAS  Google Scholar 

  13. Letai, A. Clin. Cancer Res. 21, 5015–5020 (2015).

    Article  CAS  Google Scholar 

  14. Fisher, R.I. et al. N. Engl. J. Med. 328, 1002–1006 (1993).

    Article  CAS  Google Scholar 

  15. Oakes, S.R. et al. Proc. Natl. Acad. Sci. USA 109, 2766–2771 (2012).

    Article  CAS  Google Scholar 

  16. Seashore-Ludlow, B. et al. Cancer Discov. 5, 1210–1223 (2015).

    Article  CAS  Google Scholar 

  17. Johnson, J.I. et al. Br. J. Cancer 84, 1424–1431 (2001).

    Article  CAS  Google Scholar 

Download references

Author information

Authors and Affiliations

  1. Department of Systems Biology, HMS LINCS Center, Laboratory of Systems Pharmacology, Harvard Medical School, Boston, Massachusetts, USA

    Marc Hafner, Mario Niepel & Peter K Sorger

Authors
  1. Marc Hafner
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Mario Niepel
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Peter K Sorger
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to Peter K Sorger.

Ethics declarations

Competing interests

The authors declare no competing financial interests.

Integrated supplementary information

Supplementary Figure 1 Effect of division time (Td) on conventional and GR metrics.

(a) Simulation of responses to two drugs inducing either partial growth inhibition (blue) or cell death (green) in slow- and fast-growing cell lines for a 72 h assay. Colored lines correspond to the cell count over time for the maximal drug effect; black lines correspond to untreated control samples; x0 is the cell count at the time of treatment; xctrl and x(c) are the cell counts in the untreated and treated samples at the end of treatment. End-point ratios (x(c)/xctrl) are reported. (b) Simulation of the effects of a hypothetical mutation on cell division time (by row) and drug sensitivity (by column) in comparison to a reference response (center plot in gray). The untreated population is shown in black and the treated population in color based on the type of response (partial growth inhibition in blue, cytostasis in red, or cell death in green). End-point relative cell counts are reported. (c) Conventional (top) and GR-based (bottom) drug dose-response curves corresponding to the conditions in (b). Equations for each method is based on the values defined in (a). Note that the GR curves for each type of response overlap. (d) Conventional and GR-based sensitivity metrics corresponding to the dose-response curves in (c).

Supplementary Figure 2 GR metrics for the gCSI dataset.

Each panel shows GRmax plotted against GR50 values across all cell lines for a given drug. For illustration purposes, GR50 values are capped at 31 μM.

Supplementary Figure 3 GR metrics and entropies for the CTRP dataset.

(a) Entropies for GR50 and GR2μM for each drug in the CTRP dataset. Size of the circle reflects the mutual information between GR50 and GR2μM. (b) Joint entropy of GR50 and GR2μM (H50+2μM) metrics plotted against the entropy of the GRAOC value (HAOC). (c) Distribution of entropies for GR50 (H50), GR2μM (H2μM), GRAOC (HAOC), and joint entropy of GR50 and GR2μM (H50+2μM). Red line marks the median value. (d) Each panel shows GR2μM plotted against GR50 values across all cell lines for drugs or drug combinations with the highest joint entropy of GR50 and GR2μM (H50+2μM). For illustration purposes GR50 values are capped at 31 μM.

Supplementary Figure 4 Correlation between the gCSI and CTRP datasets.

(a) Spearman's correlation values between the two datasets based on conventional AUC (x-axis) or GRAOC (y-axis) for the nine drugs shared between the datasets. Mean increase in correlation and p-value of a t-test are reported. Note that gemcitabine (blue cross) is not included in the statistics because it was used for calibration. (b) Each pair of panels shows the AUC (left) or GRAOC (right) values calculated on the CTRP dataset versus the values on the gCSI dataset for all cell lines for each drug shared between the datasets. Each pair of plots is a different drug; Spearman's correlation is displayed at the top of each plot.

Supplementary information

Supplementary Figures

Supplementary Figures 1–4 and Supplementary Methods (PDF 96 kb)

Rights and permissions

Reprints and Permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Hafner, M., Niepel, M. & Sorger, P. Alternative drug sensitivity metrics improve preclinical cancer pharmacogenomics. Nat Biotechnol 35, 500–502 (2017). https://doi.org/10.1038/nbt.3882

Download citation

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1038/nbt.3882

This article is cited by

Search

Advanced search

Quick links

Nature Briefing: Cancer

Sign up for the Nature Briefing: Cancer newsletter — what matters in cancer research, free to your inbox weekly.

Get what matters in cancer research, free to your inbox weekly. Sign up for Nature Briefing: Cancer