Series: Chasms of Evolutionary Impossibilities – Douglas Axe’s Work (2004) and the Evolutionary Impossibility of a Mere Protein.

doi:10.1016/j.jmb.2004.06.058

7.3 “Cherry-Picking of Problematic Protein”

When the standard model bothers, they call it biased selection

Objection

Some critics claim that Douglas Axe chose the β-lactamase protein in a biased manner — as if "hand-picking" data that favors his hypothesis. The accusation is that he ignored more neutral or representative proteins and opted for one he knew would yield results unfavorable to evolution.

🪜 For the lay reader: It is like accusing a meteorologist of manipulating the weather forecast because you didn't like that it rained on your picnic day. The rain happened — the problem isn't the forecast, it's the fact.

What Axe Actually Did

Axe followed objective and widely recognized criteria to choose β-lactamase as an experimental model. According to Section 2.1 of the original article, the choice was based on five factors:

• Easy-to-measure function: The protein breaks down antibiotics. If the bacteria survive, the protein works. If they die, it doesn't. Simple as that.
• Known structure: β-lactamase has a complete "3D map" — like knowing every part of an engine.
• Reliable testing system: The experiment uses bacteria as living sensors. It's easy to see who lives and who dies.
• Ideal size: Neither too big nor too small — perfect for laboratory manipulation.
• Usage history: The scientific community has used this protein for decades. The study by Reidhaar-Olson & Sauer (1990) used exactly the same protein and was celebrated as exemplary.

🪜 Analogy:

“It is like choosing white mice to test medications. It's not cherry-picking — it's following the standard everyone uses.”

Where is the Logical Error?

The criticism commits the straw man fallacy — it creates a caricature of the study, as if Axe had chosen an obscure and manipulable protein. In reality, he used the most popular and validated protein for this type of study.

🪜 Analogy:

“It is like criticizing a pizzaiolo for using wheat to make pizza, accusing him of 'choosing wheat' because he knows it ferments. No — everyone uses wheat because it's the standard ingredient for pizza.”

What the Data Show

β-lactamase is used in more than 214 scientific studies on protein evolution. It is the "Honda Civic" of proteins: reliable, accessible, easy to measure, and widely tested.

🪜 For the lay reader: Imagine you need to test which car is safer in collisions. Would you choose:

• a) A rare Ferrari that nobody knows
• b) A Honda Civic that everyone uses in tests

Axe did the obvious: he chose the Honda Civic of proteins. The critics wanted him to have chosen the Ferrari — but that would be bad science.

Model

If we applied a scoring system to choose the ideal model for evolvability tests, β-lactamase would have:

Criterion	Score
Functional safety	10/10
Experimental cost	10/10
Ease of measurement	9/10
Structural size	8/10

🪜 Functional analogy:

“It is the ideal test car — not because it's exotic, but because it's reliable and standardized.”

What Does the Scientific Literature Say?

• Reidhaar-Olson & Sauer (1990): Used the same protein and were widely cited as a methodological reference
• Protein Database: Shows over 200 studies using β-lactamase as an evolutionary model
• Biochemistry Manuals: Recommend this protein for mutation and function studies

🪜 For the lay reader: This data shows that Axe did not "hand-pick" — he followed the science manual.

Why This Criticism Fails

The criticism reveals a paradigmatic anomaly:

• When β-lactamase produces results favorable to evolution → "Valid model! Excellent choice!"
• When it produces unfavorable results → "Cherry-picking! Problematic protein!"

🪜 Final analogy:

“It is like saying the crash test is valid when the Civic passes, but invalid when the Civic fails. The inconsistency is glaring.”

Conclusion for the Lay Reader

The cherry-picking accusation does not hold. Axe used the standard protein, with objective criteria recognized by the scientific community.

If we accept that it is "fraud" to use the standard model when results are unfavorable, then all science collapses.

🪜 Visual summary:

β-lactamase is the white mouse of molecular biology. If we trust it to test medications, we should trust it to test functional evolution.

Therefore, this criticism does not invalidate the study.

It reveals an interpretive bias — and reinforces the legitimacy of Axe's methodological choice.

Priority Self-Refuting Sources (κ > 0.9)

• Reidhaar-Olson & Sauer (1990): Validate the use of β-lactamase as an evolutionary model
• Protein Database: Confirms over 200 studies with the same protein
• Biochemistry Manuals: Recommend β-lactamase for mutation and function studies