>>>>> "joe" == Joe Landman <landman at scientificappliance.com> writes: joe> What I am really getting after is whether or not there are real joe> possibilities for building a business out of providing software and joe> service for informatics computing. It seems to me that there are many joe> business models that simply do not work here (ASP is a great example). joe> I am wondering if being an independent bioinformatics software vendor is joe> a workable scenario. I think there are, especially with respect to helping to support open-source projects. I'll point out 2 examples which are very real, and while not "molecular sequence analysis" (which is what most people here seem to think of), are somewhat relevant (one being bioinformatics, the other being "general tools"). #1 - A project which I'm helping with as developer and release manager, BioConductor (http://www.bioconductor.org/), which is a open source package to analyze expression arrays, consists of packages from a number of (computational) statisticians working on statistical issues (normalization, differentiation, classification/clustering of profiles and expression phenotypes). The lead coordinator is currently trying to find commercial entities looking at providing support (because of the reputation of the individuals involved, there are groups interested in the software, and would be willing to pay for support -- which we don't really want to provide, except in that it helps us extend/fix/improve the tools we are working on). #2 - the core developers for the language at the base, R (www.r-project.org) are very interested in helping people to provide training and support. After all, while the software itself is GPL'd, there are a large number of companies (about half the major pharmas) who use it in one way or another. joe> Another way to put the question is, where is the value that a company or joe> a group would be willing to pay for. Is it in the software side? The joe> services? The implementation? The management? Where is the value, and joe> are people willing to pay for it. I think the value is in the consulting and training. Tasks which are truly simple for me to accomplish, require this for others. As with all powerful software, training is a serious issue (if you've ever attempted to use Rosetta Resolver, for example :-)... And I would suggest that there is a market if the costs can be made manageable -- i.e. a group I work with has established that given the cost of a simple piece of commercial expression array software is on the order of $3k-$6k, and labs will be willing to consider a service model to get the science done. If the software is free, the training is an issue. Even if the software is not free, training is STILL an issue. best, -tony -- A.J. Rossini Rsrch. Asst. Prof. of Biostatistics U. of Washington Biostatistics rossini at u.washington.edu FHCRC/SCHARP/HIV Vaccine Trials Net rossini at scharp.org -------------- http://software.biostat.washington.edu/ ---------------- FHCRC: M-W: 206-667-7025 (fax=4812)|Voicemail is pretty sketchy/use Email UW: Th: 206-543-1044 (fax=3286)|Change last 4 digits of phone to FAX (my friday location is usually completely unpredictable.)