In the run up to the trial of a case in which we’re arguing that the B6C3F1 mouse ain’t a man and 1,3 butadiene ain’t a human carcinogen just because it causes cancer in the B6C3F1 mouse, out comes "Mice Fall Short as Test Subjects for Humans’ Deadly Ills" by Gina Kolata of the NYTimes. And it’s a bombshell. Kolata reports on the paper "Genomic responses in mouse models poorly mimic human inflammatory diseases" and its central finding that immune responses in the mouse, including those related to heart disease and cancer, are no more closely correlated with human responses to the same stimuli than the roll of a pair of dice. It’s the long-sought explanation as to why e.g.

"every one of nearly 150 drugs tested at a huge expense in patients with sepsis has failed. The drug tests all were based on studies in mice. And mice, it turns out, can have something that looks like sepsis in humans, but is very different from the condition in humans."

Good stuff, though not all that surprising if you’ve been following the sad tale of the development of drugs that cure cancer in mice yet have no effect in humans. And it doesn’t mean that all scientific studies done on mice are worthless. Far from it. The ability to produce for example so-called knockout mice, rodents lacking a particular gene required to make a particular protein, allows an otherwise forbidden glimpse into the workings of the tiny chemical factories that we call cells. Nevertheless, the study does shatter the assumption that those little factories in mice run just like their counterparts in humans.

However, that’s not the end of the story. If you read the whole thing you’ll find this:

The study’s investigators tried for more than a year to publish their paper, which showed that there was no relationship between the genetic responses of mice and those of humans. They submitted it to the publications Science and Nature, hoping to reach a wide audience. It was rejected from both.

… reviewers did not point out scientific errors. Instead, [one of the authors] said, “the most common response was, ‘It has to be wrong. I don’t know why it is wrong, but it has to be wrong’ ”

which leads to our final point. Daubert’s peer review factor was intended to serve as an independent indicator of reliability. The Court assumed that disinterested scientists on the lookout for bad science served as gatekeepers of the journals through which "scientific knowledge" was disseminated. Perhaps when there were far fewer journals and far fewer academics desperate to be published peer reviewers served such a function. Nowadays they too often serve the status quo - barring from publication the sort of disruptive findings that would discomfit the guild they serve. Thus, if we’re not careful, does Daubert risk being effectively transmuted, at least in part, into Frye – i.e. a test of general acceptance rather than a test of sound science.