OXFORD--BioInform recently sat down for a conversation with Tony Marchington, CEO of leading bioinformatics software company Oxford Molecular Group. Founded in 1989, Oxford Molecular recently completed the acquisition of Genetics Computer Group Inc. (GCG), the developer of one of the most commonly used software packages in bioinformatics, and of the software business of MLR Automation, whose software is used in managing automated high-throughput screening data. The company's other products include MacVector, GeneWorks, GENESEQ, TOPKAT, CAChe, Cameleon, and RS3 Discovery. Marchington shared with BioInform his thoughts on his company's most recent acquisitions, important trends he sees in bioinformatics software, and his plans for the future at Oxford Molecular.
BioInform: What characteristics of GCG and MLR made them acquisition targets for Oxford Molecular?
Marchington: Building companies like ours, which aim to offer a world-class set of products and services, the most important ingredient by far is people. Any acquisition we make is always essentially about people's skills and experiences that they bring to the organization. Both GCG and MLR have exceptionally highly qualified people in that respect. GCG is a company where there are 23 people, but I would name particularly John Devereux, who in many ways is a sort of father of computer bioinformatics; he was one of the first guys ever to put a gene sequence into a computer. GCG was established within the University of Wisconsin in 1982, they've been around longer than anybody. Alongside John is Maggie Smith, who's been with John all through that period and is extremely experienced in bioinformatics. A more recent acquisition is Steve Smith, also a very talented guy. Over at MLR we have the founder there, Chip Allee, who is very talented and I'm very excited about what he will bring to the software solution needs of our customers.
Turning beyond people to the individual companies, I think GCG stood for many years as the premier company in bioinformatics. The Wisconsin Package is probably the most widely used of any, particularly in academia. It has over 600 sites and, I believe, 30,000 users worldwide. The technology will complement what we have already very successfully. Our new-generation solution for bioinformatics, Omiga, by September will be a fully networked, enterprise-wide solution for bioinformatics across PC- and Unix-based networks. In time the Wisconsin Package will slide into that very effectively. So GCG had the technology and also the user base, which will be a whole new set of customers when we introduce Omiga.
The programming team at GCG is also very talented. The company will remain in Wisconsin long-term and we hope to expand that team to be our development center for bioinformatics in the United States, as we've expanded our programming capability with another acquisition in Baltimore, which was the Chemical Informatics Division of PSI International that we acquired last year. As that's become our programming center in the States for small-molecule databases, GCG will become the programming center for bioinformatics.
GCG was, in many ways, an ideal match for Oxford Molecular. The other thing that struck me when I looked at the company was that they've achieved consistently growing sales over a number of years without, really, any sales force. There's a tremendous unexploited potential. These guys still operate out of a very modest set of offices in Madison, and we can take their practice and export it through Oxford Molecular's worldwide network, and that's going to be very promising.
As to MLR and the product we've acquired there, we've been working with that company for over a year now and they've developed an interface that sits on top of our RS3 technology, which is proving to be one of the key components in the customization involving RS3, particularly for high-throughput screening applications. It was obvious that the relationship between the two companies was getting closer and closer and then at one stage we said why don't you become part of the group, which they did. That is going to become a very important piece of technology to have, and we have plans for further interface products along the same line. The key interaction there will be between MLR, just outside San Francisco, and Baltimore, where they have the RS3 main development.
BioInform: Are you actively looking at any other acquisition targets at this point?
Marchington: Yes, always.
BioInform: There's been a big emphasis recently in bioinformatics on object-oriented technologies. How will Oxford Molecular use those technologies, and how do you see the whole trend developing?
Marchington: The way we're developing our software, a key characteristic is an extremely modular approach. Modularity brings a number of important advantages. You can have the same look and feel across a very wide range of functionality. All the pieces that share the same functionality are used in different applications, so for example if you use a spreadsheet application in small molecules, then that same spreadsheet application is used in bioinformatics. Although it's a very different program you have the same look and feel and the same key functionality. That's one important reason; the second is, once you develop one module, or one object if you like, you don't have to develop it twice. Quality goes up enormously. You've got something which is highly modular, highly object-oriented, and having got a piece of code that you know is good, it's there for many applications to come.
Next, modularity makes developing new applications much easier, because they often involve three or four modules, typically one or two that you've already got. You can take something that might have been developed for a small-molecule application or a molecular modeling application and you take out three or four pieces of core functionality and you place it among the other modules and you've got a bioinformatics application. We're finding that to be a tremendously powerful way of developing code.
And finally, in commercial terms, it means you can actually launch products in a modular fashion. So although you say to customers this what we're looking to present two years down the line, a fully integrated, enterprise-wide network solution, you can actually introduce that piecemeal, with a little piece here and a little piece there. But they sit on a fundamental architecture that allows them to synergize with each other as the solution builds. And we've been very aware at Oxford Molecular for four or five years now that that's the way we should develop these things.
We've also been aware that in the absence of any real established standards for architecture in this area--though CORBA seems to be emerging as that standard--that we should ensure that we engineer in a way that, as the standard emerges, we are ready to adopt it. For that reason we are CORBA-compliant in this company, we are Netscape-compliant, and obviously Microsoft-compliant. It's just the modern way of programming, and we saw that as a big shift that came towards the end of the eighties and as we entered the nineties, as Oxford Molecular was started. We see it as one of the key advantages we have over many competitors, because many of them have been around a lot longer than we have, but their products are now becoming very, very dated in terms of the programming technology they use. By entering this field really significantly in the nineties, rather than being one that's been around for most of the eighties, we think we've got a tremendous advantage.
BioInform: Has Oxford Molecular joined the effort to promote the IDL standard?
Marchington: We very much support the CORBA initiative. We are not formally part of it, but we would very much like to be, and we're making a pitch to give whatever support we can and also ensure that it's part of the way that we're developing here. Any successful architectural standard has to evolve. We have our own architecture, called COMS manager, but we've made it entirely CORBA-compliant, and that's maybe the best of both worlds.
We offer customers a choice of exactly how they want to string these things together. It's absolutely key that to do effective drug discovery you need to hop among very disparate scientific methods, transferring data, transferring ideas, cross-method analysis. Find me a gene that has this sequence in it, find me all the proteins that are related to that gene, find me all the crystal structures of those proteins and any small molecules that might have been acting on those proteins, design me a library, do me a literature search which has any of these protein names or any of this functionality in the abstract, or whatever. And what better way to really be able to do that than the technology that has already been developed for the net. We're finding that a lot of that stuff is transferring straight into the solutions we're building for our customers.
BioInform: What about your goal of reaching profitability? Will it happen this year?
Marchington: I can't give you a profit forecast, but if you look at our P&L report, revenues are doubling every year and that is set to continue. We don't see overheads growing at anywhere near the same rate as revenues and margins. We've got our interims coming up pretty soon, and I think the market is expecting a slight improvement--not a massive improvement--over this time last year at the half-year, but a very large improvement in the second half. And the company is very comfortable with those expectations.
Coming in the next issue of BioInform: in part two of our exclusive interview with Tony Marchington the Oxford Molecular CEO discusses the move toward platform independence in bioinformatics software, his company's approach to collaboration in software development, the trend towards creating software for less expert users, and his strategy for finding the best bioinformatics talent.