A steadily growing group of companies is using a variety of technologies to help shorten the path from platform to product company by creating improved versions of known and, in some instances, already marketed therapeutic proteins. Companies like Neose, Genencor, Applied Molecular Evolution and Maxygen, hope that by starting out with molecules already known to possess therapeutic properties, they can reduce the risks normally associated with drug discovery and development. The technologies employed by these second-generation protein players vary widely: from new methods, like directed evolution, to refinements of older ones such as glycosylation modification and PEGylation. And their business models run the gamut from a pure human therapeutic focus to the use of protein engineering in agriculture, industrial chemicals, as well as health care. But the fact remains for each of these companies that, whatever the improved odds for success afforded by second-generation protein work, they're still in a very risky business where more than half of their programs will fail. And, at least when it comes to the small universe of already marketed drugs, there will be plenty of competition among the companies. Still, there may be plenty of lucrative service opportunities to go around, as biotechs desperate for pipeline diversification, and with money to spend, look for ways to fully exploit the potential of their protein drugs.