With a bioinformatics base, even improv biology labs can get the work done
By Brad Stenger
The business of bioinformatics software changed around 1998, 1999. That’s when investors called companies’ bluff, asking, “If the software’s so great, why don’t you build some lab space, bring in some biologists, and start doing something really valuable?”
Integrated Genomics is a Chicago-based company whose founding technologists created WIT (“What is there?”), comparative biology software, at Argonne National Lab and the University of Chicago. Development kicked into high gear at IG by the mid-’90s, leading to its current ERGO platform. As software progressed, the company caught onto experimental biology and now thrives.
What’s telling about biological research that springs from software, at least the way it’s practiced at IG, is that where the approach matches the strengths of the software, results can be quite impressive — even when the labs are makeshift.
Case in point: IG focuses on acquiring an in-depth understanding of a large number of microbes, bacteria, and simple eukaryotes such as fungi. Its core software, ERGO, is a tool for executing comparative analysis, highlighting functional and sequential differences between genomes. ERGO also generates metabolic reconstructions based on chromosomal sequences and metabolic modules.
Realizing the full benefit of the software has led IG to develop distinct front- and back-end processes and facilities. Their presence in the fully occupied, multi-tenant Chicago Technology Park means that as staff grows, workgroups disperse — they’re currently housed in three different buildings.
The front end of IG’s process consists largely of a sequencing factory. Four ABI Prism 3700s and 20 Amersham MegaBaces run 24/7, dedicated mostly to sequencing microbial genomes that will serve IG or its customers or that add to the public domain.
Facilities making up the process back end — the software development, computational analysis, and wet-lab biology — are all cramped quarters. Without any relevance to the automated front end, the bioinformatics analysis work gets done in places fairly removed from the sequencing factory and the nearby 10-person microarray group. Almost everyone I spoke with said, “Yeah, we need our own building.”
The infotech department lives in a densely packed office building 100 yards away from the main lab facility. Despite the distance, the connection with lab staff is sufficient to keep feedback continuously flowing to hone the in-house software development. That in turn keeps Andrei Osterman, VP of research and development, busy and enthusiastic. “What we’re doing here is at the very [leading] edge of science,” he says, noting that he probes truly interesting scientific problems rather than focusing on what will be most profitable.
Just looking at the lab spaces, though, might leave you feeling skeptical of such cutting-edge claims. They all have a do-it-yourself quality that no lab planner or architect would take credit for. Shelves that could have come from Home Depot sit atop nearly every lab bench. My guess for why: the company has shuffled furniture around so often that reattaching original shelves became a futile exercise.
President Robert Haselkorn wouldn’t change a thing. “We spend our money on our people,” he says. “And you saw all those machines.” It’s true. Even though Haselkorn was referring to the sequencers, every software developer I saw worked simultaneously on a workstation and a laptop.
Support for the quality of IG’s work includes ongoing collaborations with top-flight academics (with presumably bigger and better facilities): synthetic biochemistry at Cornell, tuberculosis research at Stanford, x-ray crystallography at University of Texas Southwest Medical Center. These provide necessary augmentation to the ongoing work in the company’s spartan labs. After seeing IG’s far-flung in-house operations where the workgroups are an eclectic blend of foreign nationals, U of Chicago grad students, hacker-quality programmers, and corporate scientists, it’s clear that the culture is one that could successfully collaborate with anyone in any corner of the globe.
The way software initiates biology at IG is pragmatic: research goals stay within the bounds of the software’s capability. The pragmatism informs facility decision-making. In the zero-sum game, money spent on top-of-the-line people and equipment leads to labs that would make a minimalist proud. But while labs are unsophisticated, collaboration is strong, a good indicator that the electronic ties binding the organization are solid.
Brad Stenger is a freelance journalist who researches human-computer interaction in computational biology at the Georgia Institute of Technology, designs bioinformatic interfaces for Yale’s Gerstein Lab, and worked as a laboratory planner for architectural firm CUH2A. Send your comments to Brad at [email protected]