Genome Technology from August 2003 contains two articles questioning the reliability of microarray results: “Guidelines on the Way?” p. 45 and “Cross Platform,” p. 46. In “Guidelines on the Way?” the FDA dives into the messy world of microarray data and tries to come to conclusions from millions of data. In “Cross Platform” Andy Brooks says, “Lab-to-lab variation is the greatest source of error.”
Why does microarray technology not follow the principles of analytical chemistry, established about 100 years ago? Two samples should be sent to labs, these labs should do analysis of differential gene expression, results should be collected and analyzed for comparability. Then we would know about variability of results, and we could implement QC, QM, accrediting, certification, norm-administration, standard work practices, performance testing, and rule-cards. Laboratories successfully participating in performance testing would be certified to do reliable analysis; all others would have to improve. If microarray technology would do its homework, it could become a quantitative analysis method.
I believe Genome Technology would be a perfect place to start a discussion about how to get reliable results from microarray experiments.
Director of the Microarray Unit
OSU Comprehensive Cancer Center
Pay for Professors?
You omitted results for salaries from academic positions in your August issue covering the salary survey responses. Our salaries are significantly lower than any figure you published in your article. Unemployed life scientists who may be applying to academic research positions should definitely be made aware of the huge discrepancy between life in the private commercial sector and life in the public academic sector. I’ve been receiving résumés for our job openings; applicants (without a background in academia) just have no clue about the realities of salaries in academic research positions.
Jennifer Steinbachs, PhD
Deputy Director, The Center for Genomics and Bioinformatics
I read the article “Who’s in the Genome Money” (August 2003) with interest. I have a question about the reporting, however. In the methodology section, you describe the various organizational segments reached by the survey including pharmaceutical/ biotechnology (38 percent of respondents), university (28 percent), government (nine percent), other (25 percent). The graphs on the following pages are titled “Median salaries in pharma and biotech by organization size.” Do these three graphs include only the 38 percent of total respondents from this category as the title states? Or are all survey respondents included? If the former, will you be reporting the breakdowns for the other sectors?
Editor’s note: Salaries of academics were left out of the August cover story due to an editing oversight. See the table on p. 17 of this issue for the data. — AB
In GT’s September issue interview with Agilent’s Chris van Ingen, the number of microarray chips that the company has sold was misstated. The company sold 20,000 chips in the last quarter and 200 scanners in the last two years.