Harvard Medical School - Dana-Farber Cancer Institute


Pharmacogenomics: Predicting drug response based on genotype profiles

In the context of precision medicine, choosing the most optimal drug or combination of drugs for a particular patient becomes more important. While assessing cancer subtypes, stages, and grades and treating a patient accordingly has been implemented in clinical practice routinely, response to therapy is still uncertain in many cases. Incorporating knowledge of genomic-drug associations into treatment decision will result in preciser medication and is hoped to improve response to therapy.

It is now being appreciated that the molecular profile determines the cancer phenotype, also affecting drug response. While one patient may benefit from a particular drug another patient with a clinically similar cancer may experience adverse side effects. Predicting how a patient responds to a given drug is subject to pharmacogenomcis/toxicogenomics, relatively new fields that combine traditional pharmacology/toxicology and genomics on large scale, with the ultimate goal of ensuring maximal efficacy and minimal side effects for a patient.

Our group seeks to contribute in understanding the role of genes and gene products in drug sensitivity, drug toxicity, and the development of drug resistance. We hereby conduct integrative and statistical data analysis on large pharmaco- and toxicogenomic datasets, many of which are available in public databases. Most of the molecular data consists of mRNA and SNP data, but more data from the human proteome and also metabolom will be available in the future. The potentially ambiguous character of a drug is of high interests. Therefore, predictive models that take this ambiguity based on biological data into account are one our main research areas.

Heat map representing the Spearman’s rank correlations between gene expression profiles of cell lines; the order of cell lines is identical in rows (CCLE) and columns (CGP). Note the high similarity of the gene-expression data between CCLE and CGP.

Inconsistency in large pharmacogenomic studies
Haibe-Kains B, El-Hachem N, Birkbak NJ, Jin AC, Beck AH, Aerts HJ, Quackenbush J
In collaboration with the Haibe-Kains lab and Quackenbush lab, the CIBL published a study in Nature comparing two large pharmacogenomic datasets.
Enhancing Reproducibility in Cancer Drug Screening: How Do We Move Forward?
Hatzis C, Bedard PL, Birkbak NJ, Beck AH, Aerts HJ, Stern DF, Shi L, Clarke R, Quackenbush J, Haibe-Kains B
This publication suggests a path forward to establish better practices and standardizations of critical steps in pharmacogenomic assays.
Importance of collection in gene set enrichment analysis of drug response in cancer cell lines
Bateman AR, El-Hachem N, Beck AH, Aerts HJ, Haibe-Kains B
The publication assessed the influence of selected gene-sets for Gene set enrichment analysis (GSEA). The results show that GSEA findings vary significantly depending on the collection chosen for analysis.
Comparison and validation of genomic predictors for anticancer drug sensitivity
Papillon-Cavanagh S, De Jay N, Hachem N, Olsen C, Bontempi G, Aerts HJ, Quackenbush J, Haibe-Kains B
In this study the use of preclinical model systems has been intensively investigated as this approach enables response to hundreds of drugs to be tested in multiple cell lines in parallel.
Back to Top