Mailing List Archive: 49091 messages
  • Home
  • Script library
  • AltME Archive
  • Mailing list
  • Articles Index
  • Site search
 

Bioinformatics

 [1/8] from: jim:bergenshaw at: 16-Sep-2002 22:08


Is anyone aware of anyone using Rebol in bioinformatics. We are considering using Rebol to develop a distributed data integration engine. Jim Shaw [jim--bergenshaw--com] -- Binary/unsupported file stripped by Listar -- -- Type: application/ms-tnef -- File: winmail.dat

 [2/8] from: sqlab:gmx at: 17-Sep-2002 7:57


Hello Jim If you count distributing data between clinical information systems, then yes. I use Rebol to tie some (HL7-)gateways. AR

 [3/8] from: reffy:ulrich at: 17-Sep-2002 6:27


That sounds interesting .. any more information on distributed data integration engine? Dick

 [4/8] from: jason:cunliffe:verizon at: 17-Sep-2002 7:59


> That sounds interesting .. any more information on distributed data
integration engine? For general reading about IT in Life Sciences I highly recommend http://www.bio-itworld.com [free!] A lot of discussion on distributed databases, component architectures, pattern matching, and much more. What Physics was to the 20th this is to to the 21st century ./Jason

 [5/8] from: greggirwin::mindspring::com at: 17-Sep-2002 10:55


Hi Jim, << Is anyone aware of anyone using Rebol in bioinformatics. We are considering using Rebol to develop a distributed data integration engine. >> I haven't seen anything, but I've got a book on using perl for bioinformatics. I think REBOL could be really great (given its list processing abilities) but it would just be a fun thing for me to tinker with (I like cellular automata and such, hence my interest) but I haven't had the time to spend on it. --Gregg

 [6/8] from: tomc:darkwing:uoregon at: 17-Sep-2002 10:44


what sort of "distributed data integration" are you interested in? genome/sequence annotation? If so maybe looking at http://biodas.org/ would be a good start for protocals and then, "apollo" at http://sourceforge.net/projects/gmod/ "Artemis" at http://www.sanger.ac.uk/Software/Artemis/ and "fmap" currently integrated into Acedb but soon to be free standing at http://www.acedb.org/ for models of apps that would such a protocal I project of this sort would be of acute interest to me. On Mon, 16 Sep 2002, Jim Shaw wrote:

 [7/8] from: jim::bergenshaw::com at: 17-Sep-2002 15:40

Re: Bioinformatics and commercial software products


We develop data mining software products for life science research labs. Our products are used to optimize complex gene sequencing, geneotyping, gene expression, biomolecular screening, and other high throughput research processes. Our customers are large, collect data from thousands of experiments a day, and utilize many different computing platforms (SUN, Linux, Windows, Compaq Alpha, IBM, etc.) The data that we mine may come from a number of different sources; relational databases (usually Oracle), XML databases and files, structured files, flat files, etc. We are planning to develop a distributed data integration engine to expedite integrating our data mining products with new customer data sources. Over time we may add the facility to integrate data from major public and proprietary data sources and provide data to products other than our. We will need to distribute the data integration task over a number of servers (and/or PC grid?) in order to complete it in the time required; usually less than four hours. Since Rebol is cross platform (we only need to run it on commercial servers and large Windows 2000 or latter desktops) and provides built in support for network protocols, it may be a good choice to develop our new product in. However, I do have some reservations. I have not found any references to major bioinformatics systems written in Rebol. Can Rebol handle the data volumes we require, 10+ GBs? There is no biorebol.org, like there is a biojava.org, a bioperl.org, a biopython.org, a bioruby.org, etc. Why not? Ruby is a newer than Rebol yet it already has bioruby.org. I have not found any references to enterprise software products written in Rebol. Why not? Can anyone point me to some? I have been playing with Rebol whenever a I have some free time over the last couple of weeks and like what I have discovered. However, I'm still not convinced that it isn't fooling to attempt to build a $25,000 software product on it. I welcome your comments. Jim

 [8/8] from: jason:cunliffe:verizon at: 17-Sep-2002 19:46


> I have not found any references to major bioinformatics systems written in > Rebol. Can Rebol handle the data volumes we require, 10+ GBs? There is no > biorebol.org, like there is a biojava.org, a bioperl.org, a biopython.org, a > bioruby.org, etc. Why not? Ruby is a newer than Rebol yet it already has > bioruby.org. I have not found any references to enterprise software > products written in Rebol. Why not? Can anyone point me to some?
Hi Jim It sounds like a terrific project.. To my knowledge no BioRebol yet. Java, Perl, Python, Ruby all have substantial roots in academia, and thus natural flow for science departments Python in particular benefited from early adopters at NASA, Lawrence Livermore National Laboratories and few other places.. A tiny of people liked what they saw, ran some tests, and then implemented bold policy which included development of very significant math/science and inter-operability modules. Java, Perl and Python all benefited from dotcom boom. Became standards, huge on-line communities, sites, books, reputation and growing population of students and professional to draw on for labor force. A rich open symbiosis of computer science culture and application domains has blossomed Ruby came late and has its largest following in Japanese Universities I believe. There are many comparisons to Python, on whose tails it rides.. http://www.thinkware.se/cgi-bin/thinki.cgi/RubyVsPython Rebol IS truly different. It is innovative, elegant, effective and unique. And waiting to be discovered and applied in major new application domain. Yours is perhaps just such a one. But Rebol still has a tiny community, is little known and runs under a much discussed proprietary license. That tends to slow its adoption, especially in academic/science dept. circles where openSource is center stage, perhaps more for reasons of perception than reality. But even tiny price tags can become bureaucratic obstacles. For a commercial product like you describe, this may also give you some strong leading advantages. On top of the implicit internet features, Rebol's external library and shell capabilities give you access to a great richness of other toolkits. REBOL's elegant 'port' model, derived from the revolutionary AmigaOS, should also be a great help to you. Likewise simplicity of installation, especially on Windows. You ask about large data volumes.. 10Gb+ Can you clarify how those volumes need to be accessed and moved? I read a very interesting article in BioIT World magazine recently, said that key issue now for Life Science software developers is they need the skills of qualified scientists more than programmers. So any platform which reduces the learning and prototyping curve/time is preferable. Rebol fits that. Database connectivity and XML handling are presumably two areas you need to consider. I recommend you contact REBOL Technologies directly, perhaps visit them to discuss your needs in person. They are brilliant and take the long view seriously, so an ambitious development like yours merits that experience and fresh thinking. I hope REBOL meets your needs. Else I suggest Python. Good Luck ./Jason ______________________________________________ Jason Cunliffe [NOMADICS: Director art+design] Tel/fax: +1 718 422-1078 [jasonic--nomadics--org] N 43:00.000' W 074:31.875' ALT:1144 ft 84 Henry Street #3C Brooklyn NY 11201 USA