Win Hide Puts South African Bioinformatics on the Map
by Ken Howard
Dan Davison tells a classic Win Hide story. Davison, associate director of bioinformatics at Bristol-Myers Squibb, was driving home one day when he suddenly heard his former post-doctorate mentee’s voice on the radio. That summer in 1991 Hide had published a paper in Nature showing, via molecular evolutionary analysis, that the guinea pig had different origins than previously believed. The discovery was getting widespread media attention. “Win got on NPR,” recalls Davison, “and along the way of talking about the discovery, he mentions what guinea pig tasted like. He said that as long as he had muscle tissue, he thought he should taste it.”
Davison chuckles. “I almost drove off the side of the road. Not many scientists would say that.”
Win Hide is a man who likes to make an impact. Most recently, that has entailed creating an internationally respected bioinformatics presence in his South African homeland. “We’re a very long way from Europe and a very long way from the understanding of our own people,” says Hide. “We want to bring technology to the Third World and also discovery to the First World.”
Before the time came to put South Africa on the genomics map, however, he honed various skills — scientist, politician, administrator, businessman — outside his country. First in the UK, where in 1981 he earned a BS in zoology at the University of Wales, and then in the US, earning a PhD in molecular genetics at Temple University and performing post-doctorate work in computational genomics at Baylor College of Medicine, the University of Texas, the Smithsonian Natural History Museum (where he dined on the guinea), and the University of Houston.
During this time he merged his biology and genetics knowledge with computational approaches to studying genes, eventually designing bioinformatics algorithms and software. His next move, to head up the bioinformatics effort at California-based MasPar Computer, led to business skills and the woman who would become his partner.
“I learned a lot about venture capital and what a company needs to be successful,” Hide says of his year at MasPar before the company was closed down. It was during work on a joint project with nearby IntelliGenetics (later acquired by Oxford Molecular Group), that Hide met Tania Broveak. “We had this joke,” says Hide. “She was software and I was hardware.”
The joke had been on Hide four years earlier at Baylor when he first spoke to Broveak over the phone. He had been doing his charming best to get himself and academic friends free IntelliGenetics software. Says Broveak, now his wife: “Win was post doc and he wanted us to give him 200 copies of free software. I laughed him off.”
First family of South African bioinformatics
Hide may have gotten the last laugh, however, or at least vindication for the logic of his request. After returning to South Africa to be with his dying father, Hide founded the South African National Bioinformatics Institute (SANBI) and, with Broveak, South Africa’s only bioinformatics company: Electric Genetics. The two entities work closely together, with Hide running SANBI and Broveak heading (with Hide as chief scientific advisor) Electric Genetics.
SANBI develops bioinformatics software in addition to performing gene discovery work and Electric Genetics further develops and then markets the software. Its business model? While charging companies for the software, all products are distributed free to academic scientists.
“Everything must be free to academics so it can be debugged. All our products are peer-reviewed,” says Hide. “If companies want to buy it, they know it’s tried and true, so it reduces marketing costs, which in South Africa we can’t afford. Plus, academics go into corporate jobs and tend to bring products with them.”
Hide had returned to South Africa at a time of racial rebuilding and got the ears of politicians interested in putting money into piecing together science programs at traditional black universities. “Just after apartheid, it was time to transform colleges into solid scientific institutions,” Hide explains. “It was a naïve endeavor, but I knew it didn’t matter. It was going to be like doing science on the moon anyway, so I went to a place where I got a lot of support.” The place was the University of the Western Cape, where in 1996 Hide established SANBI with a government grant.
“South Africa is a long way away from everywhere,” agrees Australian Tim Littlejohn, the founder and chief scientific officer of eBioinformatics and member of the board of SANBI. “I understand the challenge of being separated from the intellectual center. This indicates Win’s a high risk-taker, and he also can reap the rewards.”
Linux junkies and smelly software
The risks may not have been as high for Hide as others, however. The example and connections of his father, who had diplomatic status in the US as director of the South African Council for Scientific and Industrial Research, certainly helped Hide’s cause. Hide also turned the lack of domestic competition and resources to his advantage. Says Broveak: “Being here is great for staffing. We can attract the best computer science people.” She adds, “By being in South Africa there is also a different level of creativity, there are no preconceptions about how systems work.”
There is also the rich resource of open-source programmers: Hide says Cape Town is home to Linux junkies who don’t want to leave the country. “They contribute heavily to open source and they are our core programmers at SANBI and Electric Genetics,” he says. “Linux is our core system so I can load it onto my laptop and work while I’m traveling, run experiments 36,000 feet over the Atlantic” reconstructing, comparing and analyzing whole genomes.
At the core of that analysis is the Hide/SANBI-developed and Electric Genetics-marketed StackPACK, a data assembly and transcript analysis system allowing clustering, alignment and analysis of high volumes of expressed sequence tags (ESTs) and the construction of partial length sequences into full-length gene sequences. Says Littlejohn: “We have the StackDB database made using Win’s software and it is being used by excess of 10,000 people worldwide. The software is important. It is the best technology to identify variations based on test errors, splicing errors and polymorphic variation. Win’s decided to be very specialized in ESTs and is among the best in the area.”
“ESTs are like books with pages torn and scattered all over; you must reconstruct the book from the pages,” explains Hide. “This is difficult to do, the technology teases out information and changes it to make it less fragmented.”
Hide’s use of pilot gene discovery technology, which he calls “bad, rank, and smelly, but better than nothing” creates a virtual gene that builds a story from the little bits. “Not a perfect story,” he admits, “but better than blank sheets of papers.” Hide says his team intends to make a virtual gene set for every organism for which it can find data.
SANBI’s research also focuses on discovery of human genes linked to genetic disease. They scored a success in 1999 with the discovery of the retinitis pigmentosa gene, which is associated with a form of hereditary blindness. Also of particular interest to Hide and his SANBI crew are the virulence genes of pathogenic organisms devastating South Africa: TB, HIV and malaria. “We have the highest rates [of the diseases] in the world so it is important that we contribute to knowledge of virulence genes to understand how the pathogens get into the body and how they survive,” says Hide. “It is also very important for me to bring South African science into the fore of the Human Genome Project.”
The Next Big Thing
To accomplish his goals, he is already looking towards the Next Big Thing. “I do not expect to be spending more than a couple years on DNA-based gene discovery,” he says. “The field is going towards functional biology and there is a need for development of biological systems-based analytical — not reductionist — tools which when put together will generate knowledge.”
The man who once mentored Hide and nearly missed a car crash over Hide’s penchant for having his guinea pig and eating it too has faith in Hide’s foresight: “He has a very good sense of what is scientifically important and ways to build tools to answer the questions so the software arrives at the same time the questions are asked,” says Davison. “This involves risk in asking the right questions and developing the right software, but he has done well with that.”
Exporting those ideas to the rest of the world means frequent travel away from the mountains surrounding his Cape Town home. A recent month’s trip took him lecturing to China, France, and the US. Reaching his audience and moving forward with research means constant contact with the “human community, passing and summing knowledge between individuals,” says Hide. This, he asserts, will require continued racking up of frequent flyer points to supplement heavy e-mail trafficking.
But Hide also sees information gathering and analysis on another scale, one conveniently removed from geographic barriers. This idea may represent the next step in his software development and one that he claims will allow the integration of the astronomical amount of data pouring in from structural biology: “Bio-mation-based technology—automation slightly artificially intelligent — will create a system within the Net which you do not need to understand,” Hide explains. “It gathers information and can turn off or on to access knowledge. It is the only way to build a system complex enough to understand complex systems, to fight fire with fire.”
To keep up with the marketplace of ideas and knowledge, Hide also recognizes the need for a continued presence on the conference circuit, where he is beginning to see himself and friends as elder statesmen. “I’m becoming first generation,” says 39-year-old Hide. “There are new people, the second generation, who are starting to look old.”
His advancing age hasn’t seemed to curb Hide’s enthusiasm for being “an extremely hyperactive bioinformatics mascot” according to his wife, and also for post-discussion drinking as both friends and Hide himself report. “I like to drink. Ask anyone at any conference. I’m the one most likely to be lying in a pool of my own problems,” he jokes.
The drinking part seems to be true and the lying down in anything but a clean hotel bed a bit of an exaggeration aimed at a reporter. Just as he made himself taste the guinea pig and then reveled in telling the tale, Hide seems to enjoy seeing how far he can go.“I say what I think. I take the piss out of everybody, including myself,” he says.