Why Big Data alone lacks the punch pharma needs

Pharma outfits have gobbled up Big Data sources on cancer genomes and transactional records, but intelligent use of data sets brings much more value to drugmakers, Guy Cavet writes for Genetic Engineering & Biotechnology News (GEN). Cavet, vice president of life sciences for Kaggle, covered how biopharma outfits have started to scratch the surface, analyzing and harnessing large amounts of data to discover, develop and market therapies.

Kaggle, where Cavet works, organizes online competitions in which researchers use data sets to come up with the best solutions to problems and compete for prizes. His organization is among a growing number of groups that are transforming the way biopharma companies tackle scientific and business conundrums with large data sets. But Big Data alone isn't enough. Technological advances have drastically sped up the aggregation of scientific data, but this has not necessarily changed the slow and unpredictable process of bringing new therapies to patients.

"With drug development costs rising and approvals declining, new approaches are sorely needed," Cavet writes. "It's too simplistic to see 'big data' as a knight in shining armor, but the intelligent use of rich data, regardless of size, has the potential to help dramatically with problems from basic research to commercial operations."

Cavet noted the following ways life sciences outfits have succeeded with intelligent use of data:

  • Analytics. Banking the data is step one. The next step involves slicing and dicing the data to derive some value from it. While not mentioned in Cavet's article, Mount Sinai School of Medicine and the software company Cloudera have been working together to develop tools for analysis of large and complex sources of patient data for research and discovery. One of the big ideas from the collaborators is to harness and analyze multiple sources of Big Data to guide and improve patient treatment.

  • Competition/crowdsourcing. Scientists can be very competitive and most seem to love a challenge. Kaggle organized a competition sponsored by German drugmaker Boehringer Ingelheim that led to a better way of predicting small molecule safety. As Cavet notes, Netflix proved in 2006 that allowing outsiders to compete with each other resulted in an improved system for recommending movies to its customers. Kaggle is doing the same for science.

  • Privacy. Whenever sensitive data are involved, security becomes a major concern. Cavet writes that there are ways for drugmakers to collaborate with outside groups on data-driven research without losing complete control of their data. In the case of Boehringer Ingelheim's Kaggle competition, the pharma group was able to share data sets with the contestants without revealing the structures and activity profiles of the small molecules.

- check out Cavet's piece in GEN