In the first of a series on jobs for the new millennium, Ayala Ochert examines bioinformatics and its dearth of skilled practitioners
One of the most important jobs for a scientist is the accurate prediction of the future. But all too often scientists themselves neglect to predict the future of their own job. Young scientists starting out on their career consider what will keep their interest for years to come, but think far less about what will keep the interest of those who will be funding their research.
The new field of bioinformatics is one that certainly has the interest of those who hold the purse strings in science, but few biologists have the unique combination of biological knowledge and computer expertise required.Those who do will find themselves very much in demand from drug companies and will probably be able to name their price.
Germinating from the aftermath of the molecular biological revolution, bioinformatics thrives on the mass of new data that it has generated. It embraces the computer processing and manipulation of molecular biology's raw data, from genome and protein sequencing, to determining the structure and function of proteins, and ultimately to the designing of new drugs on the basis of this detailed information.
"In truth, molecular biology is a soft form of information processing," says Alan Mills, formerly of Birkbeck College's x-ray crystallography department. "Genes stored in the DNA program are played out in the molecular machinery of the cell, realising the organism in its full richness." So powerful has this metaphor been, that it is now transforming the way that molecular biology is being done. Less and less happens at the lab bench, more and more at the computer screen, and even those molecular biologists not directly involved in bioinformatics now rely heavily on the software it produces.
While bioinformatics is making itself indispensable to biologists, it is the links with the Human Genome Project and the pharmaceutical industry that will secure its future for years to come. The ambitious plan to map all the genes in the human body has the backing, financial as well as spiritual, of the international biological community and through it the United States, the European Union and Japan. And with robotic DNA-sequencing machines currently churning out around 1.5 million bases (units of DNA) every day, there is enough data to feed an army of researchers. And despite some controversy surrounding the Human Genome Project, the momentum it has already generated will last well into the next century. It is bioinformatics that is needed to decipher that information, particularly for use in so-called rational drug design, making it invaluable to the pharmaceutical industry. The US company Smith-Kline Beecham has already earmarked $2 billion over the next five years to develop the technology, and other drug companies are doing the same. "Traditionally, it takes 10,000 compounds before you get one drug on the market," says Peter Murray-Rust of Nottingham University's pharmacology department. "A better understanding of the target, at the molecular level, makes it easier to predict what they're doing."
Using bioinformatics techniques to work out the structure and function of proteins, drug discovery researchers can make better guesses about which compounds are likely to work as drugs. "In the area of HIV, a lot of companies have come up with active compounds by knowing the structure of HIV protease or transcriptase," adds Murray-Rust. Despite this rosy future,bioinformatics is suffering from a serious shortage of skilled staff. "These people are like hens' teeth," says Mills. "They are like top-flight IT professionals who can pretty much name their price." Though a career in bioinformatics could become very lucrative, it is not for the faint-hearted, demanding real aptitude for computers along with a genuine understanding of the biology.
To get those skills, you need go no further than your computer screen, as there are several virtual courses on the world wide web, like those offered by the Virtual School of Molecular Sciences run by Murray-Rust. He says that it is possible to come to bioinformatics with little or no knowledge of biology, as long as you have a strong computer background. "But," says Mills, "you have to understand the biological meaning. Computers are important, but there has to be a brain engaged somewhere. Mindless trawling of databases will only get you so far."
While it is true that the generation of data in this area is almost completely mechanised, working in bioinformatics remains intellectually challenging. Although it would appear to be the epitome of reductionist science, breaking all life down into bits and bytes, it is still possible to be creative. As Murray-Rust explains, "Knowing the phone directory doesn't tell you how a city works."