From: Bob Leif To: cyto-inbox At the last ISAC, Ken Buetow from NCI/NIH spoke on "Combining Bioinformatics and Genomics to gain insight into the Etiology of Cancer. He stressed the importance of standardization in bioinformatics. The central role of XML in this process was described in his slide NCICB "virtual" organism. The lecture and slides are available at the ISAC website. I must admit in my case, Dr. Buetow was preaching to the choir, since I had a poster on CytometryML. I submitted the paper to Cytometry three months ago. It is on its third referee. Change almost inevitably leads to tension. My ISAC poster and the CytometryML schemas and XML documents can be found at www.newportinstruments.com Analytical Cytology (Cytomics) data must be able to interface with other software with as little effort as possible. Since FCS is an example of what Dr. Buetow elegantly described, "Different groups that are all spread out over the entire scientific landscape generating little stovepipes of pieces of information that are very difficult to bring together. Exacerbating that problem is these different communities all have different little languages." Since FCS is a different little language, software interfaces have to be created. This significantly increases the cost of entering the market. A standard based on XML has the great advantage of interacting with products from: Microsoft, SUN, IBM, Adobe etc. Since the datatypes in CytometryML are based on DICOM (Digital Imaging and Communication in Medicine), we can benefit from the work involved in the creation of DICOM and interface with DICOM. A significant advantage of CytometryML is that the datatypes for the large binary objects are simple and in many cases industry standards. In the case of image data, if the FCS design is followed, the data is buried inside of an all encompassing file. If it is kept separate as TIF or JPEG files, then commercially available software, such as Photoshop including its plug-ins can be used. -----Original Message----- From: J.Paul Robinson [mailto:jpr@flowcyt.cyto.purdue.edu] Sent: Thursday, September 05, 2002 10:20 PM To: cyto-inbox Subject: Re: Sweatshops of biotechnology in response to Akos Szilvasi comments... ------------------------------------------------------------------------ ------------------------------- Well, I cant resist this one. I am going to do a Mario... There are several basic issues here that have to be addressed. One is the outdated, backward, slow and insane (did I miss anyone?) data analysis techniques we use and another the rather ridiculous way we deal with samples. I have to say that over 10 years ago, I wrote what I thought was a dammed good series of grants to NIH to deal with smart data processing and sample analysis in flow cytometry. The gurus of the field (many of whom still claim to be gurus) at the time totaly poo-hoo-ed the ideas and I have the most wonderful collection of quotes saying how unnecessary automated data analysis, classification schemes, and clever data analysis was, and how flow cytometry does not need any additional analysis techiques..... In fact one of the leaders for the NIH in AIDS told me personally that the ideas were stupid!!!! - 2 years ago he called me and asked me to resubmit my grants "because they had reached a crisis" and I can't repeat what I said to him in return.... The manufacturers have never made a serious attempt at addressing this problem. They all have a totally meiopic view of what these instruments should be used for - they have totally focussed on this single sample, single tube concept - they design their instruments to do this - they pretty much ignore what happens to the data - this of course is one of the most significant problems today (it always has been, but its really bad today) There are many ways to approach these problmes, but those who sit on the holy thrones of NIH committees, need to open their minds to possibilities othere than the ones they think of....... of course, I am not telling anyone anything am I ? Paul Robinson ----------------------------------------- On 4 Sep 2002, at 16:27, Akos Szilvasi wrote: > We are a rather high volume biotech flow lab. I sent out a desperate > letter over a year ago about the lack of automated cytometers. The > situation is much worse this year. The volume is increasing (we generate > up to 2.6 GB data a month - just analyzes, without the sorter files). Most > of that "manually" on the Caliburs. A genuine, authentic sweatshop like > the ones in Asia making sneakers or other garment. > > The flow cytometry labs are the bottleneck of biotech research. We slow > down the progress by not being able to handle enough samples. The > manufacturers admit the problem, acknowledge the need but in response new > 9 color MANUAL cytometers come to the market with the promise of a FUTURE > automated sample handling extension as a teaser. > > The sad and disappointing aspect is that the whole biotech and other > research is automated. The technology is out there. Only we have no access > to it because no one bothers to adopt it (if they can not invent such > devices). > > How do you run 500+ sample experiments? > > > Regards, Akos > > (PS: This is a 10+ years old request. ) J.Paul Robinson, PhD PH:(765)4940757 Professor of Immunopharmacology Professor of Biomedical Engineering Purdue University FAX:(765)4940517 EMAIL:jpr@flowcyt.cyto.purdue.edu WEB: http://www.cyto.purdue.edu
This archive was generated by hypermail 2b29 : Sun Jan 05 2003 - 19:26:21 EST