Why would more psi-BLAST iterations find PR.C if even one iteration can’t find even the most remote homologs?
Just asking.
Ohno criticized Okada’s hypothesis as being unrealistic because it required so many point mutations, yet Ohno’s hypothesis required an essentially random amino acid sequence arising from a frameshift to instantly form a stable, functional, and specific enzyme. Ohno had no direct evidence that the hypothetical PR.C protein even existed, and his frameshift mutation was purely hypothetical. Yet Ohno put forward his hypothesis so forcefully that readers accepted his model as if it were history, and his paper continues to be cited as if the hypothetical frameshift mutation was actually an observed fact.
Many readers have not realized that Ohno’s 1984 claims were not supported by any type of evidence – his model was entirely speculative. Ohno presented his assertions very forcefully, as if they were facts. It seems that many readers of that paper got the impression that Ohno actually had observational evidence for the existence of his specified precursor protein and his specified frameshift mutation.
Careful reading shows that Ohno’s proposed precursor protein and his proposed frameshift mutation were only inferred. Therefore, at that time Ohno did not even have a testable hypothesis.
Ohno claimed that a random string of amino acids could reasonably be expected to give rise to a specific, functional, beneficial, and stable enzyme.
(Actually, this one is just a flat-out lie.)
In the beginning of Ohno’s 1984 paper, he describes why he developed his nylonase hypothesis. He had no particular interest in nylonases – he wanted to show that frameshift mutations might be a powerful way to evolve novel and beneficial proteins instantly. The falsification of Ohno’s nylonase hypothesis does not in itself prove that beneficial novel proteins might not possibly arise via frame shift mutations. However, the spectacular failure of Ohno’s nylonase model puts his broader thesis in very serious doubt.
5 Likes
Yes, but it returns the unshifted nylB gene!
No, you say the opposite, that those examples need to be revisited in light of your results.
To get this published, to get this past me as a reviewer, you need to make this point clear. Perhaps (haven’t established this yet) the nylonase example is wrong, but other examples might have stronger evidence backing them. Say that directly and you’ll fully avoid the reviewer objection.
2 Likes
What about if you lower cutoff? Did you verify that the frameshift target was in the database? Did you measure directly the blast score of that target?
1 Like
When I run BLASTP with the first 20 or so residues of Ohno’s PR.C, I get a hit in the refseq database.
Any conclusions drawn from the minimal BLAST searches you did, @stcordova, have to be considered to be very, very tentative.
OTOH, there is enough of a literature that argues against Ohno that your minimal study really adds nothing to the field. (In case you are still collecting reasons why your paper was rejected.)
1 Like
Each iteration it is able to find more and more remote homologs. You have to run it with enough iterations that it converges to a stable answer. One is not enough. Might as well do a standard blast search.
2 Likes
I think those are good criticisms and worth considering.
Now that I’ve humored a LOT of smaller issues, the big issue is, “is Ohno wrong about NylB”? You don’t have to refer to my paper to make that determination, but here is good evidence from Yomo that we referenced:
In 1992, in response to Kanagawa’s discovery of p-NylB, Yomo et al. co-authored a paper with Urabe and Okada, to put forward a third competing hypothesis regarding the origin of NylB. Yomo et al. argued that Kinoshita’s f-NylB and Kanagawa’s p-NylB homologs descended from a common ancestor that existed about 140 million years ago.[5] Yomo et al. wrote: “The distance between P-nylB and FnylB (or F-NylB′) is much larger than between F-nylB and F-NylB′. The time divergence of F-nylB and PnylB is estimated to be at least 1.4 x 108
years… Therefore, most of the amino acid substitutions from the ancestor of the nylB gene family to its descendants of today might have occurred before the beginning of nylon manufacture.”
That puts deep suspicion about Ohno regarding NylB, and we have independently confirmed this through another method.
CONVERSELY, why should Ohno be right? You can see his reasoning here:
Perhaps the most guarded language is “serious doubt”, but anyway John Harshman a couple years ago emboldened the writing of this article because his opinion was Ohno 1984 was wrong.
If you address and rectify the small and large issues, I’ll positively support your claims. Both the small and large issues are important if that is the goal.
I did, and that failed to find PR.C.
I viewed Psi-BLAST as the most relaxed and permissive search. I don’t think more iterations make it more relaxed, Art suggested more iterations would actually be more constricting (less positives, false ones too), not less constricting (more positives).
I can mention plain vanilla BLAST failed to find PR.C and a more relaxed psi-BLAST search failed to find PR.C, but it’s pointless to run another iteration of psi-BLAST to look for PR.C if it doesn’t even find a single hit on the first run.
I can clarify that the purpose of the psi-BLAST was not to find NylB remote homologs (which are many), but rather PR.C homologs.
Where did I ever say that?
1 Like
My understanding is that Ohno was wrong. I wrote a little about it on the Biologos forum months ago. The paper that shows that he was wrong was in JBC and I pasted it below. Credit to @Agauger for correctly critiquing the nylonase frameshift claim.
Of course, this means that showing this again is not of any value. And no matter what, the reasoning that if one guy was wrong about one frameshift, then everyone else is wrong about the other ones is… utterly ridiculous. I haven’t examined the paper’s methods.
6 Likes
Well thank you for the input. I appreciate this criticism because the above is extraneous to the fundamental issue : “Did PR.C actually exist.”
If PR.C existed, there is no point in writing the paper. I would appreciate your view on whether PR.C existed.
1 Like
Thanks Steve.
My main concern is that we were technically correct. Whether an editor considers what we are saying as sufficiently novel is another issue.
We could only point out that Ohno 1984 continues to be cited since Negoro’s publication and that a more forceful statement might be in order.
Art and Josh and Dave criticisms I think are fair. I have not much objection to the issues they raise, and I think they can be remedied.
1 Like
It is possible our critiques can be addressed. However they also mean we do not think the study is currently technically sound.
1 Like
FWIW, we referenced the paper Steve just shared as evidence for our claim.
You are completely right about that.
1 Like
Another point - in this age of automated annotation, genes that may actually encode PR.C may well be annotated as nylonases, and their entries in protein databases may be based on the translations that yield nylonases. So while it may be (and probably is) the case that the hypothetical PR.C gene never existed, the BLAST analysis you did, @stcordova, still leaves a bit of room for doubt.
2 Likes
Thanks Art. I think that can be remedied, albeit with some work because the Exact sequence that we’re looking for was specified by Ohno. I recall I did the BLASTN search to show this, but I didn’t include in the paper. It will rule out the single nucleotide Ohno claims existed both for NylB and other potential homologs.
FWIW, I did notice this about PR.C:
The gene coding for the NylB protein was contained in a segment of DNA Okada et al. called RS-IIA. [3] It is worth mentioning that it appears Ohno mislabeled Okada’s RS-IIA as R-IIA in his paper. [3,4] Also, it appears Ohno made either a typo in transcription or failed to clearly account for the creation of a premature stop codon in construction of his PR.C from the RS-IIA sequence. Okada’s paper and GenBank indicate that the end of Ohno’s PR.C (derived from RS-IIA) should be “GCGGCGTGA,” not “GCGGCTGA” as was the case in Ohno’s paper. Given that Okada’s paper was the source of the actual sequence data, with Ohno’s work deriving from that paper, the error must be Ohno’s and not Okada’s.
In light of that, it was easy to be skeptical that PR.C ever existed at all. But for the sake of completeness, I’ll consider looking into tying the loose end you point out.
@sfmatheson you think the paper was technically sound?