ABRF Microarray Group Seeks Submissions for Expanded Study of Chip-to-Chip Variation in Affymetrix Arrays
Seeking to gather a larger and more varied dataset for its study of variability in Affymetrix arrays, the Microarray Research Group of the Association of Biomolecular Resource Facilities (ABRF) has decided to post submission instructions on the ABRF website for any interested party.
“We’re going to expand the study to include anybody that’s running Affymetrix technology,” said Andrew Brooks of the University of Rochester Medical Center, who is leading this effort. “The data will then be published in a peer-reviewed journal, and will help us develop the standards on sources of variability that will help people in analyzing data.”
Brooks added, emphatically, that the submission parameters allow all participants to submit data, in the form of the cell (.CEL) files, without revealing any specific and confidential results. “All we need to know is chip lot number and tissue type,” he said. The other information, including the species, array type, and other information is contained within a cell file. The information about the source of the array will not be included, nor will the information about specific treatments applied to the samples, Brooks said.
The ABRF site link (http://www.abrf.org) will jump to an FTP site at the University of Rochester Functional Genomics Center, which Brooks heads. The instruction for submission of cell files will be on the ABRF site, beginning May 1st. Researchers will be given between six to eight weeks to submit data, and the microarray research group plans to roll out the first database and statistical analysis results by the end of the summer.
After a person submits a cell file to the site, the data will be cleansed and analyzed by the microarray research group, then put on the microarray research group’s website, with the permission of the submitter. Data that does not meet the “minimum performance criteria,” such as that from cross-species hybridizations, will be discarded.
Brooks and other members of the group will look at the housekeeping genes and the 3’-5’ ratios in different tissues.
The group has obtained commitments from two different organizations for over 1,000 arrays, Brooks said.
Xeotron Selects Applied MEMS to Make its High-Density Arrays
High-density microarray startup Xeotron of Houston said last week that it had selected Applied MEMS of Stafford, Texas, to manufacture its microfluidic microarrays. The company said it chose Applied MEMS, a subsidiary of seismic instrumentation maker Input/Output, because of its automated facility capable of producing microfluidic arrays.
“Applied MEMS not only provides a sophisticated manufacturing foundry that offers the specific MEMS production our biochips require, but also provides the complementary services such as custom packaging and product testing, that will assist us as we near commercialization of our first biochip products,” stated Xeotron chairman and CEO Martin Lindenberg. “Allying ourselves with a partner who can offer a complete package of services provides us with an efficient, integrated solution to our wafer manufacturing needs.”
Xeotron’s technology is based on in situ oligonucleotide synthesis coupled with digital light and photosensitive chemistry in a microfluidic chip.
In February, the company received a $3 million grant from the US Defense Advanced Research Projects Agency to develop a system for multi-strand oligonucleotide synthesis. Under the grant, Xeotron hopes to shorten the process of oligo synthesis to under 12 hours, making it possible to synthesize large segments in roughly 24 hours.
RECOMB REPORT: How Much is Enough?
Microarrays: How Many Do You Need? was the title of a paper presented by Alexander Xien, a doctoral candidate at the bioinformatics group of the Fraunhofer Gesellschaft’s Institute for Algorithms and Scientific Computing in Sankt Augustin, Germany, at Recomb 2002, the sixth annual conference on computational molecular biology in Washington, DC last weekend.
Xien, whose coauthors included his professors Thomas Lengauer and Ralf Zimmer, found that at least 15 samples in each class are required to generate reliable results from microarrays. Measuring more samples was preferable to replicating measurements of the same sample, he concluded. Additionally, he said that values should not be removed in data analysis, and that non-parametric tests are better than parametric tests.
A summary of the paper, which has not yet been peer-reviewed, is available at http://cartan.gmd.de/~zien/classize/.
Nature Biotech Paper: Roll On Down the Microarray
Molecular Staging has published a paper in the April issue of Nature Biotechnology describing the use of its Rolling Circle Amplification for on-chip amplification of signals on protein microarrays.
The authors describe a functional antibody chip that can detect 75 different low-abundance human cytokines, and another 51-feature dendritic cell cytokine chip, in the paper, “Multiplexed Protein Profiling on Microarrays using Rolling Circle Amplification (RCAT).”
The success of this technology shows that multiplexed protein analysis can be both technically and economically feasible, the authors claim. “RCA-amplified protein arrays allowed, for the first time, [greater than] 50 members of a family of proteins to be measured simultaneously without compromise of biologically relevant sensitivity,” they wrote.
Molecular Staging, of New Haven, Conn., sees RCAT as a potential replacement for PCR in some areas. Potential applications include point-mutation detection and infectious-agent diagnosis through direct detection of viral DNA or RNA.
Report: Affymetrix to be Eclipsed as Functional Genomics Reaches $2B by 2007
Looking through its crystal ball, market research firm Front Line Strategic Consulting has predicted that Affymetrix will be left in the dust within five years, as Applied Biosystems and Gene Logic develop technologies more useful to the process of drug discovery.
The report generated by the Foster City, Calif.-based consultancy further projected the market for functional genomics technologies to double, from $1 billion this year to more than $2 billion by 2007 — a conservative estimate compared to the Frost & Sullivan report issued in 2001 that predicted the microarray market would top $3 billion by 2004.
The report specifies that technologies like microarrays, gene expression, and knockout systems will expand at a compound annual growth rate of 15 percent in coming years. It also predicts a shift from target identification to target validation that will drive a 22 percent annual growth rate in non-array expression technology. This non-array based market, according to the report, will top $500 million by 2007.
According to the report, the proliferation of these technologies could be slowed by a feature that has made them so successfu l— the volume of data they are able to generate. The high volume of functional genomics data may create a glut in drug discovery, causing pharma and biotech companies to focus more on technologies that can help in downstream validation.
Other analysts of microarray technology have reported, in contrast to the conclusions of this report, that the technology can actually be useful in relieving target validation data gluts, as it becomes incorporated into toxicogenomics and early-stage clinical trials.