Main Idea

DrugTargetInspector combines mutli-omics data from a tumor sample with a priori knowledge on biological processes and pharmacological and medical databases to characterize a tumor and hence to support clinicians in their treatment-decision-making process.


Cancer is a complex class of diseases that are caused by an interplay of genetic and environmental factors and that can be characterized by a common set of features, known as the Hallmarks of Cancer. These features describe altered pathogenic processes and mechanisms caused by mutations in the genome of the cancer cells. These alterations usually induce, amongst others, high proliferation rates, failure of DNA repair mechanisms, and high mutation rates, leading to clonal evolution of tumor cells and thus to a high genotypic and phenotypic diversity between tumors and even within tumors.

Typical treatment options for cancer include surgery, radiation and chemotherapy. The latter can be subdivided into two categories: non-targeted and targeted chemotherapy. Classical chemotherapeutic drugs like cisplatin or temozolomide attack all rapidly dividing cells in the body. This low specificity leads to severe side effects as e.g. also cells of the immune system are battered. Targeted chemotherapy, on the other hand, tries to utilize characteristics of the tumor by specifically attacking target molecules that play a central role in pathways dysregulated in the tumor. Due to the heterogeneity of tumors and the fact that cancerous clonal evolution can rapidly induce drug resistance, the treatment of malignant tumors is still a grand challenge. The positive fact that the number of chemotherapeutic agents is steadily growing (currently there are more than 200 FDA-approved anticancer drugs) renders the search for an optimal treatment even more difficult, in particular, if a combination therapy is required.

In order to determine an optimal treatment for a given tumor, an in-depth characterization of the tumor’s genetic and phenotypic makeup provides an ideal basis for decision-making. With the advent of biotechnological high-throughput methods, the genetic and molecular characteristics of tumors can be measured at relatively low costs, making genotypic and phenotypic cancer profiling available to clinicians to support diagnosis, prognosis, and treatment selection in the near future. These big, complex and noisy -omics data sets require the availability of easy-to-use, yet powerful bioinformatic tools that are able to integrate different -omics data types, to extract the most relevant information, and to visualize the results in a clear manner that facilitates further analysis and interpretation.


To support oncologists and clinicians in the treatment decision-making process, we have developed DrugTargetInspector (DTI), an interactive assistance tool that provides rich functionality for the integrative analysis of tumor-specific -omics datasets. To reveal the characteristics of a given tumor, DTI analyzes and integrates genomic, transcriptomic, and proteomic datasets, where genomic data sets provide information on genetic alterations (mutations). These mutations can affect the eligibility of therapy options in different ways. On the one side, driver mutations induce the deregulation of certain processes and pathways and, hence, the knowledge of such mutations and the induced pathway activities is important for decisions on (targeted) therapies. On the other side, mutations can make therapies ineffective or reduce their efficacy (e.g. drugs targeting EGFR are ineffective if an activating KRAS mutation exists in the tumor). To account for these dependencies, mutation data can be uploaded to DTI and based on the GDSC database, DTI annotates given mutations with their pharmacogenomic effects across a large panel of drugs.

Transcriptomic and proteomic datasets can be used to identify deregulated signaling pathways and processes. To this end, DTI performs enrichment analyses on a large set of pathways derived from a variety of databases. Based on these pathway activities and the corresponding information on mutations, putative target pathways, drug targets and their corresponding drugs can be identified (e.g. PTEN deletion or loss of function activates the PIK/AKT/MTOR pathway, interesting drug targets could then be mTORC1 or the upstream receptor HER2 to be targeted by e.g. (a combination of) everolimus and trastuzumab). DTI also determines if these drug targets are deregulated and offers functionality to determine their effect on downstream processes. To this end, DTI performs a subgraph analysis, which reveals the most deregulated subnetwork rooted in a drug target of interest. The subnetwork is visualized along with its corresponding gene expression data, which allows for a visual assessment of how the downstream molecules might be influenced by the root node.

In summary, DTI provides a powerful integrated tool suite for cancer therapy stratification by providing in-depth analyses of tumor -omics datasets and a characterization of various aspects of dysregulation in the tumor, making DTI a valuable addition to existing clinical decision-support systems.


  • Future X Healthcare Scientific Excellence Award (roche.de) (2017)
  • Ausgezeichneter Ort im Land der Ideen (land-der-ideen.de) (2018)