[@BPSDB] Andrew Schlafly, owner of Conservapedia - the homosexuality-obsessed site whose edit history and block list provide the answer to the question "Is a wiki better than a blog for expressing your personal opinion to an audience" - has come back for another round of the Lenski Affair. Not content with the answers he got from evolutionary biologist Lenski about perceived irregularities in Lenski's paper, Schlafly is taking the fight to the journal which published it - PNAS. Oh yes, that's right, he's sending them a letter (da-da DAAAAA). The text is below, between the dotted lines, following which I'll give a commentary.

- - - - - - - - - - - - - - - - - - - -

Draft of PNAS Letters Response from Conservapedia

Title:

Identification of flaws in the following paper published in PNAS: Blount ZD, Borland CZ, and Lenski RE, "Historical contingency and the evolution of a key innovation in an experimental population of Escherichia coli," 105 PNAS 23, pp. 7899–7906 (June 10, 2008).

Author:

Andrew Schlafly, B.S.E., J.D.

Author Affiliations:

www.conservapedia.com, teacher of pre-college students

Text:

The following flaws in this PNAS paper negate its claim that E. Coli bacteria evolved through a beneficial mutation:[1]

1. Figure 3 depicts an "historical contingency" hypothesis around the 31,000th generation, but the abstract states that mutations "arose by 20,000 generations." The paper fails to admit that the Third Experiment disproved the contingency hypothesis depicted in Figure 3.
2. Both hypotheses propose fixed mutation rates, but the failure of mutations to increase with sample size disproves this. If the authors claim that it is inappropriate to compare for scale the Second and Third Experiments to each other and to the First Experiment, then it was also an error to treat them similarly statistically.
3. The paper incorrectly applied a Monte Carlo resampling test to exclude the null hypothesis for rarely occurring events. The Third Experiment results are consistent with the null hypothesis, contrary to the paper's claim.
4. It was error to include generations of the E. coli already known to contain trace Cit+ variants. The highly improbable occurrence of four Cit+ variants from the 32,000th generation in the Second Experiment suggests an origin from undetected, pre-existing Cit+ variants.
5. The Third Experiment was erroneously combined with the other two experiments based on outcome rather than sample size, thereby yielding a false claim of overall statistical significance.

The underlying data for this publicly (NSF) funded research have not been publicly released, despite requests for such release and NSF policy that "data collected with public funds belong in the public domain."[2]

cc:

Randy Schekman, Editor-in-Chief, PNAS, University of California at Berkeley (by email and postal mail)
New Scientist (by fax - 0171 261 6464)
Rep. Brian Baird, Chairman of the Subcommittee on Research and Science Education of the U.S. House Committee on Science and Technology (by postal mail)
Judicial Watch (by email)

References

1. ? Detail is at http://www.conservapedia.com/Flaws_in_Richard_Lenski_Study and its talk page.
2. ? http://www.nsf.gov/sbe/ses/common/archive.jsp

Comments

The word count for the above letter is precisely the PNAS limit of 250 for the Text section, excluding the cc: list. The foregoing letter is to be sent by postal mail, return receipt requested, to PNAS, 500 Fifth Street, NW, NAS 340, Washington, DC 20001, by email to pnas@nas.edu , and by posting it in its feedback form at http://www.pnas.org/feedback .

- - - - - - - - - - - - - - - - - - - -

If you're used to dealing with peer-reviewed research, the letter comes across as a bit naive. Most real academics wouldn't title the letter using confrontational language like "Flaws in...", but with "Response to...". Also, "affiliations" is for listing the places you work for, not your personal blog and occupation, which amusingly he doesn't name an institution for. The funny part thought is the CC list... Judicial Watch?! New Scientist?! And the Democrat Brian Baird?! And god help the editor of PNAS, who will be receiving no less than 5 copies of the same document for some strange reason. At any rate, suffice to say that the letter stands out as an amateur job even before we get to the actual content.

So what of the criticisms? Schlafly points to five specific errors in the paper, and then complains about missing data. I'll take them all in turn, with the caveat that I am by no means an expert in this particular part of biology (although I do have experience in peer-review). I'll take the five errors first.

1. Figure 3 depicts an "historical contingency" hypothesis around the 31,000th generation, but the abstract states that mutations "arose by 20,000 generations." The paper fails to admit that the Third Experiment disproved the contingency hypothesis depicted in Figure 3

This is a basic comprehension problem. Figure 3 doesn't depict any hypothesis based on the 31,000th generation - it's just an illustrative figure. Lenski's team were looking at one of two scenarios. If the ability for e. coli to eat citrate evolved due to a rare mutation, then the probability of the mutation occurring would remain roughly constant through-out the experiment. If, on the other hand, the history of the population made the mutation more likely to occur, then as the experiment progressed there would come a point when the probability of it would suddenly jump. The graph in Figure 3 simply describes those two scenarios. The generation in which the jump occurs is irrelevent. It is very clear that Schlafly simply hasn't read the paper in detail, or that if he has he has failed to understand it.

2.Both hypotheses propose fixed mutation rates, but the failure of mutations to increase with sample size disproves this. If the authors claim that it is inappropriate to compare for scale the Second and Third Experiments to each other and to the First Experiment, then it was also an error to treat them similarly statistically.

What Andy is referring to is that the third experiment was larger in scale than the first and second, yet the mutation rate didn't increase with sample size as expected. What he fails to mention is that the authors address this in the paper, highlighting the anomalous result and linking it to other changes in the third experiment. In terms of "treating them similarly statistically", it's a vague comment, and it's hard to understand what Schlafly means.

3. The paper incorrectly applied a Monte Carlo resampling test to exclude the null hypothesis for rarely occurring events. The Third Experiment results are consistent with the null hypothesis, contrary to the paper's claim.

Quite why Andy has decided that the use of Monte Carlo is incorrect is beyond me, and the third experiment simply wasn't consistent with the null hypothesis, disproving it with 92% certainty.

4. It was error to include generations of the E. coli already known to contain trace Cit+ variants. The highly improbable occurrence of four Cit+ variants from the 32,000th generation in the Second Experiment suggests an origin from undetected, pre-existing Cit+ variants.

Schlafly doesn't appear to realize just how precise modern microbiology tools are. It's not like Lenski and his assistants are chucking handfuls of cells around. The equipment available to modern microbiologists can count populations cell-by-cell with extreme reliability. There were no Cit+ variants in the samples, and Schlafly's assumption that there must have been is based entirely on his own prejudice that it "must" be wrong, and his apparently lack of understanding of the field.

5. The Third Experiment was erroneously combined with the other two experiments based on outcome rather than sample size, thereby yielding a false claim of overall statistical significance.

Much like point3, this is a random claim about the methodology being "wrong". There is nothing wrong with combining samples of different sizes in science - in fact I have to do it a lot myself. There are accepted statistical techniques for doing it (in this case the "Z-transformation" method), and if Andy has an issue with those techniques he should explain why.

Finally we come on to the claim about witheld data.

The underlying data for this publicly (NSF) funded research have not been publicly released, despite requests for such release and NSF policy that "data collected with public funds belong in the public domain."

This is well covered by Lenski's replies, so I'll just link to them here, but it seems that reading is not Andy's strongest subject.

How will PNAS respond to the letter? With any luck, they'll ignore it. The man is an internet crank, and irrelevent to the point that not even fellow Young Earth Creationists seem to take him seriously. Plenty of genuine researchers would like to have their views heard in the letters page of PNAS, and to give space to a random idiot would be disrespectful to all those genuine people who want to contribute to scientific debate.

And so the letter will be ignored, or a stock response sent. The cognitive dissonance in Andy's brain arising from the rejection of what he perceives to be valid complaints, will result in him branding PNAS part of some vast liberal conspiracy suppressing the truth about this "flawed" paper.

And so Schlafly will continue to rant in his empty, echoing chamber, watched with amusement by the handful of curious observers clustered around the other side of the thick glass pane of ignorance and fear that separates him from the rest of the human race, until eventually even they become bored, and the words fade into darkness.

Subscribe to similar posts from a range of bloggers at the "Blogging on Pseudo-Scientific Douchebags" Aggregator. Click here for an RSS feed.