Dr. Paul von Bünau is Managing Director of idalab, a Berlin-based agency for data science founded in 2003, specialising in machine learning, artificial intelligence, mathematical modeling and data strategy. idalab advises a wide range of clients from banking, mobility, biotech and the digital sector. A mathematician and computer scientist by training, Paul von Bünau holds a Ph.D. in statistical data analysis from TU Berlin.
Accelerating early stage drug development through rapid target discovery and triage using large scale text mining
An emerging biotechnology company located in Boston, MA and Berlin, Germany, funded by leading US/EU life science investors, is developing a platform to identify new allosteric (non-active site) approaches for important biological targets. Multiple open data sources exist for validating human disease targets in this way, but these data need to be triaged and prioritised based on clear upfront hypotheses. Together with idalab, our client analysed ~20,000 human targets, mining information from 27 million scientific publications within a timeframe of 3-6 months. Off-the-shelf software solutions were poorly aligned with the needs of the company’s unique approach. idalab developed a completely tailored solution using self-learning artificial intelligence (AI). A natural-language processing (NLP) algorithm enabled extensive text mining and thus the creation of proprietary data taxonomies for the client, focused around the company’s vision. Additionally, integrating and linking datasets from various domains enabled hidden insights to be uncovered from the data. The prioritized datasets contributed to the company nominating its initial drug discovery programs that are now yielding important drug-like molecules.
- 2:45 pm- 4:15 pm: REGIONAL SESSION II: DIGITAL HEALTH & MEDTECH
Location: Stage 2