Issue Date: February 16, 2015
First, Verify Your Research
Regarding “Confronting Irreproducibility,” I appreciate and agree with many of the points made (C&EN, Dec. 15, 2014, page 28). There is, however, much more. Biology has largely become chemistry over the past 50 years, and much of chemistry research embraces biology. Multiple problems accrue from the organizational silos and the way research is funded.
My analytical chemistry friends have much interest in getting access to “meaningful samples” to test their measurement ideas and clinically validate them. My biological friends have less experience with the rigors of obtaining good numbers using validated methods.
In the academic world, the incentives are aligned with innovation over validation. We obsessively want more publications and more chapters for our Ph.D. dissertations. We pedal faster and faster in the face of a tough funding climate that does not support careful, transdisciplinary work.
Biology graduate students use instruments they have never been trained to use and don’t understand. Chemistry graduate students examine biological samples collected in ways that alter their composition. Many students are unaware whether their commercial pipettes actually deliver 10 µL, and they don’t check them. That pH meters are not to be used with mixtures of alcohols and water was well-known in 1970 but not today. Some report data with five significant figures when the standard used was at best 95% pure.
Medicinal chemists will supply a compound dissolved in ethanol for a biologist to dose in a rat. Analytical chemists will receive blood samples and be unaware that how they were collected and the time of day they were collected can make a significant difference. Glucose meters will show three digits when the second digit is uncertain.
Be careful out there. Verify first, and then trust, not the reverse. Perhaps scientific papers should carry a phrase such as, “This method has not been validated as fit for purpose; use at your own risk.” Even if you have the claimed transgenic mouse or cell line, it’s still easier to be wrong than right.
Peter T. Kissinger
West Lafayette, Ind.
- Chemical & Engineering News
- ISSN 0009-2347
- Copyright © American Chemical Society