Selected Publication


Cancer: Subpopulations based on differential drug response

Segmentation of a population based on pharmacological patterns of response discovers subpopulations with differential response. a Dose response curves of two or more drugs are measured across a population of just over 1000 cancer cell lines. b The population is segmented into distinct and homogeneous subpopulations based on their response to multiple drugs. When comparing two drugs, subpopulations can be categorised based on their mean log(IC50s): sensitive to both drugs (orange), sensitive to drug A but not drug B (green), sensitive to drug B but not drug A (blue), and resistant to both drugs (grey). c Segmentation results for a BRAF inhibitor (SB590885) and a MEK inhibitor (CI-1040). Tree nodes contain the number of cell lines and are coloured based upon their category of response. Significance testing of 735 cancer functional events reveals subpopulations enriched for BRAF and KRAS mutations. (source: npj Systems Biology and Applications)

Two drugs, even with the same target, rarely have the same potency across all cancer patients - so how do we objectively select the right patients to treat with each drug?

In a study published in the journal ‘npj Systems Biology and Applications’, an international effort led by Michael P. Menden (Research Group Leader at the ICB of Helmholtz Zentrum München) and Dennis Wang (Lecturer and Group Leader in Genomic Medicine at the University of Sheffield, UK) in collaboration with Constellation Analytics, LLC. (USA), developed a machine learning approach called SEABED to identify groups of individuals from a population who respond distinctly to a portfolio of targeted therapies. SEABED stands for SEgmentation And Biomarker Enrichment of Differential treatment response.

Keshava & Toh et al. used SEABED to systematically compare 327 pairs of anti-cancer drugs across a panel of >800 cancer cells and 30 cancer types. SEABED identified groups of cells where one drug was more effective than the other. Interestingly, groups responding differently to the pairs of drugs could be treated more effectively when both drugs are given together as a combination, which can be in the presence or absence of drug synergy. This approach enables systematic explorations of personalised medicine to reveal biomarkers and drug combination opportunities.


Original publication:

Keshava N*, Toh TS*, Yuan H, Yang B, Menden MP#, & Wang D#. Defining subpopulations of differential drug response to reveal novel target populations. npj Systems Biology and Applications. 5, 1-11 (2019).

*co-first authors

#co-corresponding authors

For further details, please refer to the article

We use cookies to improve your experience on our Website. We need cookies to continually improve our services, enable certain features, and when we embed third-party services or content, such as the Vimeo video player or Twitter feeds. In such cases, information may also be transferred to third parties. By using our website, you agree to the use of cookies. We use different types of cookies. You can personalize your cookie settings here:

Show detail settings
Please find more information in our privacy statement.

There you may also change your settings later.