Open (source) for business

In a previous life, I worked for a bioinformatics research group that was building and maintaining a large repository of molecular interaction data, as well as several other data warehouses and bioinformatics tools. The group provided this information free-of-charge to anyone willing to visit its web or FTP sites. Likewise, we provided open-access to the code for its data analysis tools, so that researchers worldwide could set up in-house versions and tailor the tools to suit specific needs.

The group tried to interest the commercial, government, and academic research communities in the various services and data that they could offer these scientists. Time and again, however, we would look through the trade and scientific literature to see yet another licensing deal between commercial bioinformatics providers, who offered similar services and data, and these very same commercial, government, and academic researchers we were trying to attract. Large sums of money being spent on information and tools that were, to a large extent, freely available. What gives?

Some of the fault for the lack of interest in open-source, open-access bioinformatics resources must surely lie with the academic and government providers. Comprised mostly of coders and bioinformaticians, the groups have little understanding of the concepts of sales and marketing and therefore tend to generate tools that are computationally excellent but visually underwhelming and not particularly user-friendly. These are resources by specialists for specialists, and the bench biologist is intimidated by the tools, preferring to spend money on something flashier and simpler to use. Think using Microsoft Office versus working in DOS.

But part of the blame also rests with the market, which generally fears you will get exactly what you pay for when you get something for free. There is also concern about the "fly-by-night" nature of academic/government projects of this type, where loss of funding can mean loss of a resource. Ironically, I would argue, the commercial bioinformatics market has suffered greater swings of fortune in the last two decades than the academic environment with only the best providers surviving today.

The open-source, open-access bioinformatics community has many parallels to Linux in its early days. From the first blush of excitement within a small computing community, Linux has seen a steady, if slow, upward progression in acceptance by the commercial and lay markets. As with Linux, the money for these bioinformatics resources will likely lay in the peripherals and secondary applications that improve their utility. And the push by organizations like PubMed and PubChem to dramatically expand the repertoire of data available publicly should also increase the global confidence in the long-term viability of these resources.

There is yet a long way to go before science is truly democratized. However tediously, we are getting there. We just need to keep our eyes (and minds) open to the opportunities.