Michael Jackson stumbled onto his pharmaceutical path thanks to a partnership in the early ’80s between Johnson & Johnson and the Scripps Research Institute, where he worked as a scientist. When J&J finally ended the relationship in favor of spinning off its own biotech research center in 1994, Jackson and several colleagues moved to the J&J center.
He’s never looked back. Now senior vice president of drug discovery for the pharmaceutical R&D division, Jackson, 43, oversees the La Jolla-based center, which has become the hub of the pharma giant’s genomics activities and attempts to integrate chemistry, biology, bioinformatics, and genomics. “It’s taken till about now” — nearly 10 years later — “to really see the productivity” of it all, he says.
In the early days, Jackson remembers, sequence data was a prime element — his team was among the early subscribers to Incyte’s database, and the lab was “generating so much sequencing data, we couldn’t handle it,” he says.
Later, microarrays became a focus. Working with the renowned Stanford arrayers, J&J researchers built their own spotters and an in-house platform; to this day, the center spots many of the chips used throughout J&J’s research labs. As informatics for the arrays turned into a bottleneck, Jackson and his crew worked with Battelle to help found OmniViz to come up with data- and literature-mining software.
Jackson, a UK native who started his scientific career studying cell transport and antigen presentation, is particularly interested in the future of arrays. Whole-genome chips don’t interest him much, though: “There’s an enormous redundancy of data here,” he says. “I’m not sure we need all the detail.”
Jackson’s goal is actually to have less on the chip but to be able to study it more intensively, for example with time studies. He also believes that some of the major players — Affymetrix, for one — aren’t focusing enough on automation. The platforms remain “too labor-intensive” to really become high-throughput, he notes, adding that J&J can make its own chips more cheaply and get better throughput with them.