Finding Full-Featured Free Software For Biologists On The Internet

Internet Author: Robert Finn Sidebars: Access to Freeware and Shareware Archives With the life sciences becoming ever more reliant on computers, biologists find themselves constantly on the lookout for useful software. While there's no shortage of commercial vendors willing to sell full-featured packages that will perform tasks such as modeling a protein or running a flow cytometer, this software can cost many hundreds or even thousands of dollars. Fortunately, an impressive array of free or low-cost software for biologists is available on the Internet, much of it written by independent academics or government scientists. Such software is often born of a specialized need, according to Harry Mangalam, director of biocomputing at the University of California, Irvine, College of Medicine. "Generally what you'll find is some biologist getting frustrated with something not being just the way he wants it [in a commercial package], and he'll write a program that addresses exactly what he wants. Then he'll think about making it available to everybody." Distribution via electronic networks has made this step a relatively straightforward one for both software authors and users. "Making it available on the Internet is the easiest, cheapest way to do it," says biologist Don Gilbert, who manages computers in the biology department at Indiana University. On his own time, Gilbert is a prolific software writer who also maintains one of the most complete repositories anywhere of freely available biological software and data - the Indiana University Archive (see accompanying list for information on connecting to this and other resources mentioned in this article). "There aren't really any good commercial channels for [freelance] software distribution. Venues for independent authors are pretty limited," Gilbert notes. Freelance science software authors generally write their programs to meet their own needs or those of other researchers in their field. Because the potential market often is a small one, they usually don't have the option of approaching a software publisher to release their work commercially. They are left with two alternatives for distribution: "freeware" or "shareware." Both involve posting the program to the Internet, where anyone can download and use it. Freeware costs its users nothing, but shareware authors ask those who use their software to voluntarily send in a usually modest fee after an evaluation period. Some authors give their software freely to users at nonprofit institutions, and ask for fees only from commercial users. "Shareware is a step between giving it away and selling it outright," remarks Gilbert. "I've heard, and I would find believable, that 1 percent or less of people who use shareware pay for it." Shareware authors have various methods of encouraging users to pay their fees. One is to leave certain desirable features out of the publicly available version, sending the full program only to paying customers. Another is to have the program display an annoying reminder to pay up every time the program is run. Gilbert himself is the author of several highly useful software tools for molecular biologists. Among them are LoopDLoop, ReadSeq, and SeqApp. LoopDLoop is a graphic editor for RNA secondary structures. Of the other two, UC-Irvine's Mangalam says, "There has been a proliferation of different database formats for genetic information. There are about 15 or 20 now. [ReadSeq is] the Rosetta Stone that interconverts them all, pretty transparently. "SeqApp was the first sequence-analysis program that had knowledge of the Internet built into it," observes Mangalam. "Instead of keeping an entire sequence database on your own hard disk, you could get these sequences from across the Internet. "In a whole lot of ways [Gilbert] has been driving the development of sequence-analysis software. At every step he's basically gone [the commercial vendors] one better. Indeed, probably one of the major phrases that the support people for the major commercial molecular biology packages have come to dread is: 'Well, Don Gilbert does it like this for free - why can't you?' " Another software standout is Entrez, distributed by the National Institutes of Health's National Center for Biotechnology Information (NCBI). NCBI produces GenBank, a database of all known nucleotide sequences accompanied by their associated proteins and MEDLINE bibliographic records. Entrez allows users to query GenBank interactively over the Internet. According to Francis Ouellette, who coordinates submissions to GenBank, Entrez users can currently search among 227,000 protein records, 414,000 nucleotide records, and 1.2 million MEDLINE records. The version of Entrez that works over the Internet is completely free, as are the searches themselves. Users without Internet access can subscribe to bimonthly updates on CD-ROM through the Government Printing Office for $76 per year. Scientists who need to do any aspect of image analysis - quantifying electrophoresis gels, for example - can investigate another NIH program, called NIH-Image. Written by NIH programmer Wayne Rasband, "NIH-Image is an astonishing, commercial-grade, image-analysis package," Mangalam says. "No Mac owner should buy any image-analysis program without checking this out first. NIH-Image is so good that the NIH has gotten a lot of flak from other software developers. Things in Windows that do the same things as NIH-Image cost on the order of $5,000 to $10,000. This is being given away for free. Actually not so much for free - we've already paid for it out of our taxes." Of all the hundreds of shareware and freeware programs for biologists, the one currently garnering the most praise is RasMol, a free package developed by programmer Roger Sayle at the research and development labs of Glaxo Wellcome plc in Greenford, Middlesex, U.K. RasMol is used to display molecular structures, and its users tend to be highly enthusiastic. Immunologist Eric Martz of the University of Massachusetts, for example, asserts, "When I discovered RasMol, first of all, I learned enormously about protein structure in a matter of a couple of weeks by staying up late every night, which I couldn't help, it was so fascinating and so beautiful. The program is designed to be very interactive. You sit there and it draws you in to exploring the molecule." Graduate student Pedro M. Coutinho of the department of chemical engineering at Iowa State University is another satisfied RasMol user. "It's quick," he reports. "It gives a very nice rendering of the structures. It can show the structures in different ways. I'm trying to dock substrates into the active site of an enzyme, and I can quickly look at the results from a run of the other programs I'm using." Martz adds: "This is such a fabulous teaching tool for displaying molecular structure in a lecture. You can rotate the structure, and this gives you a wonderful sense of three-dimensional relationships, which you can't appreciate from fixed pictures in textbooks." One excellent source of reliable information and advice on biological software is the Usenet newsgroup bionet.software. Those who already have a pretty good idea of what they're looking for can use the standard Internet searching tools archie (best if one knows an exact or almost exact filename) or WAIS (Wide Area Information Service, for searches of text databases), or World Wide Web searching tools such as Lycos (R. Finn, The Scientist, April 3, 1995, page 17). In addition, some volunteers have already gathered this material in conveniently organized centralized repositories. Don Gilbert's Indiana University Archive, for example, contains copies of programs for molecular biology that he and many others have written, and it's available via ftp, gopher, and the World Wide Web. A great deal of similar molecular biology software is archived at the University of Houston file server called Gene- Server and at a file server at Japan's National Institute of Genetics in Mishima, Shizuoka. Two other repositories are notable for being particularly well organized: the European Bioinformatics Institute's "BioCatalog" (based at Cambridge University), and Iowa graduate student Pedro Coutinho's World Wide Web page called "Pedro's BioMolecular Research Tools." Molecular biologists have led the way in software development, and the aforementioned sites concentrate on this area of biology almost exclusively. There are a few sites, however, that archive software for other fields. The medical school at the University of Texas, Houston, maintains a site containing a great deal of biomedical software; the University of Florida, Gainesville, archives software for analysis and visualization of nuclear magnetic resonance data; and Eric Martz at the University of Massachusetts archives software for flow cytometry, some of which he himself wrote. Although a great deal of free software performs the tasks for which it was written admirably, there are a few pitfalls for potential users to consider. Says Harry Mangalam: "With shareware and freeware, you can spend a lot of time just trying to evaluate whether it's worthwhile. In the end, you may find that you've spent a week discovering that this thing is not worthwhile. You should have just spent X number of dollars and gotten [a commercial version of] this program." Don Gilbert adds, "I think the major reason people use commercial software is that it's complete. It's hard to assemble a package from free programs that will do everything that a commercial package will do in as consistent and well-supported a manner." While commercial software vendors can assemble a team of professional programmers to write, debug, and provide telephone support for a product, almost all freeware and shareware represent the efforts of a single individual. A software author frequently won't have time to answer users' questions or to solve their problems, especially if the program gets widely distributed and becomes popular. Gilbert notes, for example, "I just have to ignore a lot of the requests for support I get because I don't have the time. It hurts me to do that." And because they are not tested as thoroughly before release, free programs can fail more frequently. "Commercial products tend to be debugged much better," says Mangalam. "In the worst case, it can be so bad that you can lose your data or crash your system. But people are pretty careful, and generally when things fail, they don't fail catastrophically." Experts warn that those who download free software also run the risk of catching computer viruses. Says Martz: "Obviously there's the possibility of a virus. I've never heard of any involved in scientific freeware that I'm aware of, and I think that's probably because the vandals who do this haven't taken an interest in [software for scientists] because it's too specialized." Gilbert agrees: "I haven't found any viruses in any of the software I make available. But I think that's something that everybody has to keep in the back of their minds. Everybody should have virus-protection software." There remains the question of why a life scientist would go to the trouble of writing a program and making it available to everyone for free. Gilbert has an answer: "Biologists have always shared their knowledge. I think those who write software are just following in the same mode." Robert Finn is a freelance science writer based in Long Beach, Calif. (The Scientist, Vol:9, #15, pg.14 , July 24, 1995) (Copyright © The Scientist, Inc.) garfield@aurora.cis.upenn.edu

Finding Full-Featured Free Software For Biologists On The Internet

Interested in reading more?

Become a Member of