Computational Biology & Digital Sciences  

Using research data to accelerate decision-making and better address patient needs 

Our approach in Computational Biology and Digital Science (CBDS) centers around the use of large datasets and data analytics tools to generate scientific insights to advance novel target discoveries, create new biomarker insights, and optimize drug modality design across all therapeutic areas in Boehringer Ingelheim. To enable analysis and insight generation from data, our teams develop solid and scalable data & infrastructure solutions, such as cloud-computing, end-to-end processing pipelines, and Machine Learning (ML) enabled FAIR (Findable, Accessible, Interoperable, and Reuseable) data solutions. Our scientists and data engineers work in close collaboration with experimental laboratories, advancing method development and generating a diverse range of multi-omics data (ranging from genomics, transcriptomics, to epigenomics), and integrating it with clinical information. Advanced algorithms applied to these data, enables us to generate key insights, which are further validated with our internal and external collaborator partners to accelerate the drug discovery process.  

Research topics in CBDS at Regional Center Vienna 

Target discovery and prioritization in oncology research 

We apply computational and data science methods to large pre-clinical (e.g. cancer cell lines collections) and clinical (e.g. cancer patients) data sets, to identify new targets with higher success rate and to accelerate target prioritization efforts. To that end, we are part of the Cancer Dependency Map Consortium (DepMap 2.0), which aims to map the landscape of cancer vulnerabilities using models and high-throughput genetic and chemical screens. In addition, we are expanding our collaborations to access real-world data (RWD) from cancer patients. This would allow us to discover alternative drug targets, which might have been missed using cell-line based models. Together with experimental validations, we believe that our efforts have the potential to bring data-driven target discovery to the next level and improve the success rate of our pipeline.  

We also aim at accelerating high-priority projects, that will directly impact our oncology pipeline. For instance, we are exploring a new territory of cancer-specific antigens using state-of-the-art computational methods and multi-omics data. One key output will be to construct a unique data resource, which would enable candidate cancer antigens to be computationally prioritized according to the characteristics of therapeutics modalities, such as T-cell engagers, cancer vaccines, and antibody-drug conjugates. 

Data exploration and expanding the patient-reach for clinical drug development candidates 

Together with oncology researchers at Boehringer Ingelheim, we constantly improve on our data and analytical capabilities to accelerate decisions making along the whole value chain of drug discovery. We evaluate large-scale loss-/gain-of-function genetic, as well as drug perturbation screens in larger collections of cell line models. These approaches generate a deeper understanding into the specific mode-of-action of our novel drug development candidates, provide analysis on drug development candidates’ drug responses, or highlight hypothesis for patient selection and pharmacodynamic biomarkers. Backed up with a deep understanding of cancer biology and real-world data, we support teams, in decision making on potential new drug combinations for development candidates, or we propose novel indication expansion opportunities for existing clinical assets. Being successful in this area, would allow us to bring our treatments to even more patients in the end. 

An interconnected data and tool landscape to accelerate data-driven drug discovery 

Our data and infrastructure strategy involves partnering with Boehringer Ingelheim’s research scientists and IT, to establish a fully connected FAIR data and governance landscape. This approach enables the handling of tera- to peta-bytes of omics and other data types, generated within the whole organization and from external sources. In addition, the teams develop state-of-the-art end-to-end processing pipelines, that allow the ingestion and integration of novel data. We aim to strategically position ML/AI solutions at critical steps of our data engineering workflows, to increase the quality of data assets. Our strategy also involves the advancement of data visualization methods and tools. Visual data exploration has the potential to reveal patterns and insights, which might be overseen in fully automated solutions. 

Collaboration is key 

Our scientists in Computational Biology and Digital Sciences partner with providers for real-world human cancer data access and availability. Besides our collaboration in the DepMap consortium led by the Broad Institute of MIT and Harvard University, our partnerships  provide access to large cancer datasets, which can strengthen our understanding of tumor biology and resistance mechanisms upon inhibitor treatments. This will lead to better patient stratification and combination treatment regimens. We also invest into a long-term collaboration with the Visual Data Science Lab (Institute of Computer Graphics, Johannes Kepler University Linz), to advance visual data exploration methods and tools to derive knowledge from complex biomedical data. These collaborations have allowed us to identify novel disease targets, create new biomarker insights, and optimize drug modality design across all therapeutic areas in Boehringer Ingelheim.