By Meredith W. Salisbury
Way back in 1995, before “www” meant much of anything to anyone, Suzanna Lewis had to release a version of FlyBase in time for the annual Drosophila meeting held that year in Atlanta. Lewis — known to most people as Suzi — and her colleague Cyrus Harmon devoted hours to the production release of what would be the Encyclopedia of Drosophila, worrying about every detail, right down to what kind of jewel cases to use for the CDs. Lewis still remembers getting to the meeting that year: “We had boxes and boxes of CDs, and gave them out to every attendee,” she says. “They could just take it home, drop it in their Mac, and they had Drosophila.”
It remains one of Lewis’ favorite accomplishments, and it was an undeniable hit. “For years I would get people coming up to me and saying, ‘I still use that,’” says Lewis, bioinformatics director of the Drosophila group at University of California, Berkeley, and pie baker extraordinaire.
That certainly wasn’t her only production effort. Mark Yandell remembers a more recent one: Now an HHMI senior scientist in Lewis’ group, he was working at Celera on the Drosophila genome when he met Lewis in 1999. He teamed up with her group to get the fly genome published, “and the whole year leading up to that publication we pulled multiple all-nighters,” he remembers. “Suzi’s the only person I’ve ever seen go into a Peet’s or a Starbucks, [order a latté], and say, ‘Can I get three additional espresso shots in that?’ She’ll down it in, like, five minutes.” (The only fly gene Lewis has named is called “Pray for elves,” a reference to the fairy-tale shoemaker whose work is done for him at night by magical helpers.)
But what really impressed Yandell was Lewis’ savvy with new ideas. “She was a real advocate of XML at a time when nobody really knew what that was. We worked all night long on how to get the Drosophila genome packaged in XML — a lot of people thought she was crazy,” he says of what later became the industry standard format.
Being on the forefront is typical Suzi, if you ask Michael Ashburner of the European Bioinformatics Institute. “To some extent her forte in research is in getting new things done,” he says.
A Chem Bio
There was no path laid out for Lewis when she headed off to college at the University of Michigan. The only clear guideline she had came from friends who warned her not to take chemistry. “So I avoided it diligently,” she recalls, until her ecology major forced one chem class on her — and she couldn’t get enough. She was too far along to switch majors to chemistry, so she went for the next closest thing: molecular biology.
Chemistry appealed to her in a way that nothing else had. “I felt like Gulliver just exploring this whole new world,” Lewis says, now 54. “Every atom had its own personality, likes and dislikes.” Her near-miss of the field was itself a valuable lesson. “The other thing that that taught me was not to listen to what other people think. What may be true for them may not be true for you.”
After Michigan came California. “I’d watched far too much television and I thought that was where things were happening,” she laughs. She had grown up visiting her great-grandmother’s historic homestead in Ohio, a plot of land given out by the government as settlers land-rushed the Midwest. In California, Lewis became part of the land rush of her own generation: she moved to Silicon Valley in the ’80s and worked at computer startups. “It was when Apple … and Microsoft [were] first getting started. Everything was up for grabs at that point.”
She spent years hopping from one upstart computer company to another — writing software to help FedEx track its packages and for oil companies to keep up with seismographs — “but I always felt bad that I had left my biology behind, which is what I really loved.” Finally, she wound up leaving to write protein-sequencing software at Genenchem, a short-lived joint venture between Genentech and Hewlett-Packard.
By 1988, looking for a shorter commute, Lewis heard about the genome project starting up at Lawrence Berkeley. She signed on, discovering how much her skills were needed. “To have both a background in molecular biology and a background in computer science — nowadays, it’s a dime a dozen, but at the time that was a pretty unusual combination.”
Enter the Fly
Not too much later, Gerry Rubin came to the lab, and brought with him the fruitfly. “I had never done any fruitfly research,” says Lewis, who was eager for the fresh challenge. “I haven’t looked back since.”
Indeed. Recognized for nearly 16 years of informatics work as the “Queen of the Fruitfly” (she vigorously argues that she doesn’t deserve the honor, ticking off several women in the fruitfly community who she believes have done more than she has), Lewis is known by scientists throughout the Drosophila community she’s so committed to. “They always seem to put the science first, whatever it takes. [They’re] very collaborative,” she says.
That group needs her just as much as she does them. She’s “just the sort of person you want to be holding a fractious, argumentative community of genome scientists together,” says Sean Eddy of Washington University.
Yandell agrees. “She creates a really good environment around her,” he says. “If you put her in a group of people with five major personality conflicts, somehow you add her to the mix and it’ll all work out OK.”
That could very well explain the success of GO, the Gene Ontology group aiming to establish a standard vocabulary for naming genes, among other things, to allow people to communicate better about organisms. “We just had this little problem that we wanted to solve,” Lewis recalls of the genesis of the group. “I swear this [started] in a hallway at a conference.” She was astonished at how well everyone worked together to move the ontology goal forward. “It’s one of the biggest proofs that if people just make the decision that we are going to cooperate, we will figure it out one way or another,” she says. With GO’s success, Lewis went on to help found the Sequence Ontology as well.
EBI’s Ashburner, a founding member of GO, has known Lewis since the early ’90s when her group worked with his to figure out AceDB and FlyBase. He describes her as “energetic and imaginative — she’s probably been a bit undervalued” because she doesn’t publish much, he says. (Lewis jokes about that: while she urges her group to publish, publish, publish, she rarely takes her own advice. “I’m not trying to build a career or make a name for myself,” says the mother of five.)
Writing code is still one of her favorite activities, and though she tries “not to assign myself anything that puts me in the critical path,” some of her most fun times at the lab involve riddling through some algorithmic puzzle.
One of her current projects is Apollo, which she’s working on in conjunction with the ENSEMBL group. “It’s almost like a debugging tool for your computational analysis,” she says. “You can look at the results and decide what is the best intron/exon structure for this gene, do alternate transcripts, adjust the splice edges of your gene.”
Far from growing obsolete, Lewis’ bioinformatics skills are being called on even more these days. Up until a few years ago, she explains, “sequencing was really the rate-limiting step. Now we’re awash in genomes. Genomes are coming out faster than we can figure out what’s on them, and all of a sudden it’s us informatics people who are in the hot seat.”