CLEMSON, SC--How 28-year-old entrepreneur Maciek Sasinowski came to start his own bioinformatics company had a lot to do with being in the right place at the right time.
Three years ago, Sasinowski, a Polish-born physicist who emigrated to the US as a teenager, landed a visiting assistant professorship in mathematical sciences at Clemson University where his wife Heather had been awarded a tenure-track position teaching statistics. His only job experience since completing his doctorate at the College of William and Mary in 1995 had been a stint as a computer-game software developer. But a semester after arriving at Clemson, intellectual curiosity and a Scientific American article spurred Sasinowski to begin hanging around a wet lab run by a molecular oncologist on campus. Then, as he explained it, one thing led to another.
In 1997, Clemson plant genetics researchers Rod Wing and Ralph Dean were awarded a $2 million National Science Foundation grant to create a genomics research facility at the university. "Ralph and Rod came to talk to the math science faculty one day because they wanted to set up a collaboration with their new institute," Sasinowski recalled.
At work in the lab, Sasinowski missed the meeting. "My wife told me it was quite funny. Whenever Ralph and Rod were talking about molecular biology, the math scientists' eyes glazed over, and when the computational scientists started talking, Ralph and Rod had no idea what they were talking about," he said.
After a few minutes of confusion, someone remembered there was a computational scientist on staff who had been working in the wet lab for half the year. "Ralph and Rod jumped out of the room and offered me a position on the spot," Sasinowski told BioInform. "They were able to make me an attractive offer and it seemed like it would be fun to develop something from scratch. I didn't know at that point about bioinformatics, or that it was such a booming field, but it was probably one of the best decisions I've made in my life."
After two years as bioinformatics director of the Clemson Genomics Institute, Sasinowski began branching out into consulting work, and eventually resigned his post to run his own software and consulting services company, the Institute for Computational Genomics, or INCOGEN, full time. Having started out targeting plant and animal genomics research operations, Sasinowski lists among his clients the Novartis Agricultural Discovery Institute in La Jolla, Calif., Paradigm Genetics in Research Triangle Park, NC, Sigma-Aldrich in St. Louis, the University of Delaware, and North Carolina State University.
He spoke with BioInform recently about his experience building a startup operation.
BioInform: What were the business opportunities you saw that led you to resign from Clemson and launch your own company?
Sasinowski: For a while I was faculty at Clemson and was running this company, but decided there was no way I could do both well.
My work for Novartis Agricultural Discovery Institute is mostly contract-based software development, but they have been very supportive of me and they were one of the reasons I was able to take the leap and resign my position. They had just moved into a new building and had no infrastructure, so I knew that, even if I couldn't find anything else, I'd be able to support myself for a while working for them. As it turned out, we have more work than we can handle.
BioInform: What has INCOGEN's growth been like?
Sasinowski: We have five people right now. It started with just me doing consulting. We keep expanding the company, but one of the things we don't want to do is grow too fast.
At Clemson we went from four faculty and some students and technicians to about 80 people in two years. When I started I was the only person doing bioinformatics and by the time I left there were 17 people in the bioinformatics group. I've been limiting INCOGEN to not let something like that happen. I don't want to keep taking on projects and hiring people simply to make the company grow as fast as possible without providing the necessary managerial infrastructure to handle a large company.
My wife writes code and algorithms for statistical analysis of all the data. She's been doing bioinformatics since the Clemson Genomics Institute started and has found that she enjoys working in this field. We also hired a secretary, and the other two employees are software developers.
Being near Clemson University, we have a pool of consultants to draw from. Even though we are only five people, we have much greater resources than that.
BioInform: Have you sought out investors?
Sasinowski: No, I didn't get investors. One of the advantages of being a consultant initially was being able to earn enough money to buy a bunch of equipment. Starting a company in bioinformatics is not like starting a wet lab where you need a million dollars in equipment. It's just computers and salaries and office space.
I'm sure I could get a fair amount of venture capital, but I don't want to give up control of the company, which is always part of the price you have to pay. There are enough projects out there that could fund us for quite a while and allow me to take on a project and hire one or two more people. But right now I'm a little afraid to do that because I don't want to grow too fast and start spreading myself too thin and not do as good a job as I should be.
BioInform: How do you go about getting the jobs you're getting?
Sasinowski: Believe it or not, we have not so far had to go out and solicit jobs. People have either found us through the web or they pick up one of our brochures at a conference.
We got the North Carolina State University job at the last Plant and Animal Genome meeting in San Diego where we started putting our brochures on tables. Somebody saw the brochure and contacted me while I was at the conference. They liked what I did at Clemson and they asked me to help them set up their bioinformatics department.
Many universities are in the same position now. They've gotten the money to do some research, they know how to handle the wet lab aspect--they are experts at molecular biology, but they know nothing or very little about how to handle bioinformatics or analysis.
BioInform: As a bioinformatics consultant, are you helping install everything from hardware to software?
Sasinowski: Yes, basically everything. I can give you our last example: There's a faculty member at Delaware named Joan Burnside who contacted me about three weeks ago. I think she found me through the web. She said they have some money to do a project and that a postdoc had been working on it for the last nine months or so, but that they weren't anywhere near where she wanted to be at this stage.
She described to me that she needed an infrastructure in place to have web access to several thousand sequences and to search and query the sequences. She already knew she needed something in a database format and she wanted some web interface to allow other people access to her site. She invited me to give a presentation.
When I go to a potential customer, the first day I learn as much as I can about what they would like. Then I prepare a little presentation to make the next day, and we sit around the table and talk about what kinds of options they have. They tell me what they would like and what they would like me to do. At that point I can pretty much tell them how much the project will cost and how long it will take. I come back and prepare a proposal to send them. They might have to run it through the department head to get the appropriate funding. If everybody is happy, they approve the project and I do whatever is required.
At Delaware, I'm going to build a computer for them. I'll install the Linux operating system and a freeware database on it, which they don't have to pay for, and put the Blast server on it so they can do their Blast searches locally. Then I'll develop a web interface that they can use to query and play with their own data and also web servers that will make their data available to the public. I'm going to build the computer from scratch here and go back to Delaware and install it, and give them a bunch of documentation describing what all the things are that we've done.
BioInform: What will be the cost and length of that job?
Sasinowski: This particular project, once we get the okay, will take less than four weeks and cost about $30,000.
I know there are tons of other companies that would charge hundreds of thousands of dollars for something like that. We know that it doesn't take that. They don't have to go out and buy Oracle or a Sybase license or a Sun workstation. There are plenty of things available for free such as Red Hat Linux, which is just as reliable as anything else out there, and the database called MySQL, which is also freely available, that they can use without having to purchase a $20,000 Oracle license.
BioInform: What sort of software has INCOGEN designed?
Sasinowski: One of our specialties is web-based data visualization. We do everything in Java. We have tools to construct and visualize physical maps. We're currently working on an image recognition program to read high-density filters.
BioInform: Does it make a big difference to your customers what hardware they're running?
Sasinowski: Of course it doesn't. A CPU is a CPU. There are processors you can get for similar amounts of money and it doesn't make that much of a difference. It's possible that some vendors might be more reliable than others, but whether you're going to buy a Silicon Graphics machine that runs Windows NT or make your own and put Windows NT on it is not that big a difference other than the marketing that's involved.
What's much more important to people is how much they have to pay and what kind of support they get if something breaks down. More and more, programs are being ported to Linux now. It's much cheaper to buy a personal computer in parts and put Linux on it, and you have just as stable a platform as you have with a Sun workstation.
The trouble is that not a lot of people know how to do it. Also, if something breaks or your hard drive gives out, you have to go order another hard drive and reinstall it. With Sun, you can buy their support contract and if something breaks they come to you and you don't have to deal with it.
BioInform: Do you offer service contracts?
Sasinowski: Yes, we have service contracts that say if something goes wrong we can fix it. Or, what I've done with North Carolina State and the people at Delaware is to point out that it's much cheaper for them to hire a senior undergraduate computer science student or a graduate student to be the systems administrator. Those people know enough about computers to know that if a hard drive gives out you just replace it and install everything and your computer is operational again. Of course, you have to have backups, and that's one of the things we make sure is in place.
--Adrienne Burke