Biomarker Analysis Tool

The Biomarker Analysis Tool enables the identification of genetic signatures of tumor cell lines responsive to a customer’s drug including the gene expression signature and the mutational signature of the drug. The goal of the genetic signature analysis is to reveal prognostic biomarkers in the early discovery phase to predict response to drug treatment.

How Does the Biomarker Analysis Tool Work?

The Biomarker Analysis Tool exploits the IC50 raw data set obtained by profiling customers compounds such as the ProLiFiler^TM – Reaction Biology’s Cell Line Panel Screening service.
The biomarker identification includes gene expression levels, somatic gene copy numbers, or genetic mutations. These biomarkers are indicative for the efficacy of a drug and may serve as predictive biomarkers in disease models and patient cohorts to stratify responders from non-responders.

Contact an expert

Meet us at your next conference

Bioinformatic biomarker analysis in a nutshell

Watch to learn how Reaction Biology uses bioinformatics to identify the genetic signature of tumor cells that are responsive to the treatment of your drug and how this can help your drug discovery project. Watch to learn how Reaction Biology uses bioinformatics to identify the genetic signature of tumor cells that are responsive to the treatment of your drug and how this can help your drug discovery project.

undefined

Additional Information about the Biomarker Analysis Tool

Case Study Nutlin-3a - sensitivity data

Case Study Nutlin-3a – sensitivity data

This case study investigates nutlin-3a which is a small molecule that occupies the p53 binding pocket of MDM2 and effectively disrupts the p53–MDM2 interaction that leads to activation of the p53 pathway.

The IC50 values of nutlin-3a were determined on 274 tumor cell lines and ranked. The tumor cell lines could be grouped as resistant, intermediated sensitive, and highly sensitive. The highest testing concentration was 30 µM with semi-log increments.

The IC50 values of nutlin-3a are shown in a scatter plot for each cell line. The tumor types were ranked for the highest median IC50. The median IC50 for all tumor types (red line) coincides at 30 µM testing concentration because of the high number of resistant tumor cell lines.

The most susceptible tumor types are leukemia AML, melanoma, leukemia ALL, Hodgkin’s lymphoma.

Case Study Nutlin-3a - Gene Expression

Case Study Nutlin-3a – Gene Expression

Example of nutlin-3a gene expression analysis. Graphical presentation of Limma test results show which genes are highly expressed in drug-resistant and drug-sensitive cell lines. MDM2 and other p53 pathway-related genes were highly expressed in tumor cell lines, which were sensitive to nutlin-3 treatment.

In the table are 10 genes listed which show high expression in tumor cell lines sensitive to nutlin-3a treatment. The results show that MDM2 expression is a predictor of sensitivity to nutlin-3a treatment.

Case Study Nutlin-3a - Mutation Analysis

Case Study Nutlin-3a – Mutation Analysis

Example of nutlin-3a mutation analysis. Nutlin-3 was tested for dose-response in a panel of tumor cell lines. These cell lines contain a total of 16,998 mutated genes. The volcano plot shows for each gene (one dot) the Wilcoxon test p-value for the correlation of the drug response in proportion to the existence of mutants. For example, p53 mutations were found mostly in tumor cell lines that did not respond to nutlin-3a. KIAA1522, on the other hand, was found the most probable mutation that exists in cell lines responding to nutlin-3a treatment.

The top three mutated genes correlated with nutlin-3a efficacy. P53 was found mutated in 10 cell lines, which responded to nutlin-3a treatment, and in 152 cell lines, which did not respond to nutlin-3a treatment. The mean IC50 value of nutlin-3a in tumor cell lines with p53 wild type was about 11.6 µM compared to 27.5 µM in cell lines with p53 mutation resulting in a difference of 16 µM between mutated and non-mutated cell lines which was used as a measure for resistance.

In a mutation position analysis, amino acid Arg 273 in TP53 was shown to be the strongest marker of resistance.

Biomarker Analysis Details

Biomarker analysis at Reaction Biology is performed using a computational tool, the Cancer Data Miner, developed by our collaborator 4HF Biotec, a bioinformatics firm specializing in cancer data mining to discover new anti-cancer drugs.

Primary data set: We will perform the ProLiFiler^TM assay for testing the anti-proliferative effects of your drug on a set of 140 human tumor cell lines for IC50 determination, also called the ‘sensitivity profile’ of your compound.

Database: Our collaborator 4HF Biotec has built a comprehensive database, including data from more than 1,800 preclinical samples comprising multiple cancer datasets that are integrated into a single platform for visualization and statistical analysis.
The data relevant for the Biomarker Analysis Tool are mutations (whole-exome sequencing), gene copy numbers, and gene expression data (Affymetrix).

Approach: The ProLiFiler^TM results are correlated with the gene expression, mutation, and somatic copy number alteration data of the tumor cell lines that were used in the ProLiFiler^TM study. A correlation analysis is performed for each cell line to determine the correlation of the drug sensitivity to each gene in regards to the expression level, mutation status, and/or somatic copy number. Watch the above video for more information.

Analysis: The Biomarker Analysis is performed with two statistical tests for each correlation; one uses the IC50 value of the test drug; and the other compares the gene data of responding and non-responding cell lines.

Deliverables:

Mutation analysis

Excel tables with raw data
Graphical presentation of the correlation of mutation status for each gene and the drug sensitivity for each cell line
Graphical presentation of the correlation of the position of each mutation and the drug sensitivity for each cell line
Optional: Presentation of the amino acid position of relevant mutations

Gene copy analysis

Excel tables with raw data
Graphical presentation of the correlation of gene copy gain and the drug sensitivity for each cell line
Graphical presentation of correlation of gene copy loss and the drug sensitivity for each cell line
For selected genes, additional plots can be generated

Gene expression analysis

Excel table with raw data
Graphical presentation of the correlation of the expression of each gene and drug sensitivity for every cell line
Graphical presentation of the cluster analysis to identify differential gene expression signatures and different response rates to the test drug
Optional: Literature search of gene function for most relevant genes

Additional analyses can be performed upon requests such as target gene analysis, pathway analysis, subtype analysis, hypothesis-driven analysis, predictive model development, and more.