Changes

:: No scientific paper should present a graphical depiction of a falsehood, without indicating that it is false.  Figure 3 suggests that the potentiating mutation occurred at the 31,000th, despite the data proving otherwise.  If the authors won't correct that error, then they're unlikely to correct any other error also.--[[User:Aschlafly|Aschlafly]] 22:22, 27 July 2008 (EDT)

:: No scientific paper should present a graphical depiction of a falsehood, without indicating that it is false.  Figure 3 suggests that the potentiating mutation occurred at the 31,000th, despite the data proving otherwise.  If the authors won't correct that error, then they're unlikely to correct any other error also.--[[User:Aschlafly|Aschlafly]] 22:22, 27 July 2008 (EDT)

:::It is not a falsehood; it is a hypothesis. The hypothesis must be formed before data is collected. They hypothesis was that a potentiating mutation occurred that increased the rate of mutation to Cit+. Prior to running the experiment, it is impossible to know when (or even whether) that potentiating mutation occurred. Given that, it is not an 'error' to show the jump at the 31,000th generation, or the 15,000th, or the 10,000th - any point picked for the figure is abitrary. The hypothesis was not that a potentiating mutation took place ''at the 31,000th generation''. For Lenski to put the graph jump at the 20,000th generation would imply that he got the data and then altered the original hypothesis to make it a perfect fit with the results - and ''that'' would be an ethical breach. The data collected suggest that the hypothesis was correct in that a potentiating mutation took place - and they further suggest that this mutation took place at about the 20k generation point. This does not disprove fig 3; it refines it. Again, you may as well object that fig 3 doesn't provide numerical values on the y-axis.--[[User:Brossa|Brossa]] 10:38, 28 July 2008 (EDT)

:::It is not a falsehood; it is a hypothesis. The hypothesis must be formed before data is collected. They hypothesis was that a potentiating mutation occurred that increased the rate of mutation to Cit+. Prior to running the experiment, it is impossible to know when (or even whether) that potentiating mutation occurred. Given that, it is not an 'error' to show the jump at the 31,000th generation, or the 15,000th, or the 10,000th - any point picked for the figure is abitrary. The hypothesis was not that a potentiating mutation took place ''at the 31,000th generation''. For Lenski to put the graph jump at the 20,000th generation would imply that he got the data and then altered the original hypothesis to make it a perfect fit with the results - and ''that'' would be an ethical breach. The data collected suggest that the hypothesis was correct in that a potentiating mutation took place - and they further suggest that this mutation took place at about the 20k generation point. This does not disprove fig 3; it refines it. Again, you may as well object that fig 3 doesn't provide numerical values on the y-axis.--[[User:Brossa|Brossa]] 10:39, 28 July 2008 (EDT)

:2)This point is where I think your best argument lies, but it should be restated along the lines of: the null hypothesis suggests a fixed mutation rate rather than historical contingency. However, the largest replay experiment did not demonstrate the expected increase in Cit+ mutations, and was consistent with the null hypothesis with a p-value of 0.0823 by the Monte Carlo test method.

:2)This point is where I think your best argument lies, but it should be restated along the lines of: the null hypothesis suggests a fixed mutation rate rather than historical contingency. However, the largest replay experiment did not demonstrate the expected increase in Cit+ mutations, and was consistent with the null hypothesis with a p-value of 0.0823 by the Monte Carlo test method.

:::The statement that trace cit+ cells inevitably existed after the 31,000th generation requires that there are no techniques to remove them from the mixed-population sample. As others have told you, these techniques do exist and are reliable. You can request that PNAS publish exactly what technique was used (most likely dilution and plating on an indicator medium). Your option then is either to attack the method (difficult, since the techniques have been around for a very long time) or the execution (which would require that you or someone else repeat the experiment with the same technique and obtain different results).

:::The statement that trace cit+ cells inevitably existed after the 31,000th generation requires that there are no techniques to remove them from the mixed-population sample. As others have told you, these techniques do exist and are reliable. You can request that PNAS publish exactly what technique was used (most likely dilution and plating on an indicator medium). Your option then is either to attack the method (difficult, since the techniques have been around for a very long time) or the execution (which would require that you or someone else repeat the experiment with the same technique and obtain different results).

:::Or, taking a different tack - the paper states that it is impossible to exlude an origin for Cit+ that is earlier than the 31,000th generation. So how many generations prior to 31,000 should have been eliminated from the replay?--[[User:Brossa|Brossa]] 10:38, 28 July 2008 (EDT)

:::Or, taking a different tack - the paper states that it is impossible to exlude an origin for Cit+ that is earlier than the 31,000th generation. So how many generations prior to 31,000 should have been eliminated from the replay?--[[User:Brossa|Brossa]] 10:39, 28 July 2008 (EDT)

:5)This point is, in my opinion, the worst of the lot, mostly because it is meaningless - what does 'combined based on outcome rather than sample size' mean?- but also because it restates the claim in #3, that there was some problem with the technique used to combine the three experiments into a final p-value. If you object to the Z-transformation method, you should state why; if you object to the result, you should be able to say what you think the result should be - with a different p-number that you calculate yourself.

:5)This point is, in my opinion, the worst of the lot, mostly because it is meaningless - what does 'combined based on outcome rather than sample size' mean?- but also because it restates the claim in #3, that there was some problem with the technique used to combine the three experiments into a final p-value. If you object to the Z-transformation method, you should state why; if you object to the result, you should be able to say what you think the result should be - with a different p-number that you calculate yourself.

::: Again, show what you think is wrong by showing statistically what should have been done and what numerical results should have been obtained. Recall the bridge analogy, if you read it.[[User:Kallium|Kallium]] 23:06, 27 July 2008 (EDT)

::: Again, show what you think is wrong by showing statistically what should have been done and what numerical results should have been obtained. Recall the bridge analogy, if you read it.[[User:Kallium|Kallium]] 23:06, 27 July 2008 (EDT)

::: The combination of samples of different sizes is standard practice in the sciences - so you must be objecting to the particular technique used. If you object to the Z-method, state your reasons in the letter. If you have a preferred method, state it in the letter and defend it. Best of all, use your preferred method and present your results. By 'the hypothesis of a mutation rate', are you referring to the calculations of  mutation rates? Because the data for those calculations were not obtained from the three replay experiments.--[[User:Brossa|Brossa]] 10:38, 28 July 2008 (EDT)

::: The combination of samples of different sizes is standard practice in the sciences - so you must be objecting to the particular technique used. If you object to the Z-method, state your reasons in the letter. If you have a preferred method, state it in the letter and defend it. Best of all, use your preferred method and present your results. By 'the hypothesis of a mutation rate', are you referring to the calculations of  mutation rates? Because the data for those calculations were not obtained from the three replay experiments.--[[User:Brossa|Brossa]] 10:39, 28 July 2008 (EDT)

I would hope that you would at least run this letter past a sympathetic scientist (someone from DI?) before you submit it, to confirm that these concerns are in a legitimate format.--[[User:Brossa|Brossa]] 20:10, 27 July 2008 (EDT)

I would hope that you would at least run this letter past a sympathetic scientist (someone from DI?) before you submit it, to confirm that these concerns are in a legitimate format.--[[User:Brossa|Brossa]] 20:10, 27 July 2008 (EDT)

::: Oh how you cling to your views! The problem with logic is that it only works with knowledge- almost all of the replies on this page have been trying to teach you something about microbiology and/or statistical analysis but you are convinced you understand it better than those who have many years of experience working with them, although you continue to misinterpret basic concepts covered in any undergraduate microbiology, genetics or statistics course. Where is the logic in ignoring the input of all those who have actual training and experience in the fields you are discussing? How is that having an "open mind"?[[User:Kallium|Kallium]] 23:06, 27 July 2008 (EDT)

::: Oh how you cling to your views! The problem with logic is that it only works with knowledge- almost all of the replies on this page have been trying to teach you something about microbiology and/or statistical analysis but you are convinced you understand it better than those who have many years of experience working with them, although you continue to misinterpret basic concepts covered in any undergraduate microbiology, genetics or statistics course. Where is the logic in ignoring the input of all those who have actual training and experience in the fields you are discussing? How is that having an "open mind"?[[User:Kallium|Kallium]] 23:06, 27 July 2008 (EDT)

::: Oh, how I cling to the views of others! Yes, which is why I have an accountant and an attorney and a doctor and a statistician and an architect - I accept that other people have greater experience and knowledge in certain areas than I do. If I want to charge my lawyer with malpractice, it behooves me to check with another lawyer to see if my complaint has merit, or if I'm even articulating the complaint correctly before I show up in a courtroom.--[[User:Brossa|Brossa]] 10:38, 28 July 2008 (EDT)

::: Oh, how I cling to the views of others! Yes, which is why I have an accountant and an attorney and a doctor and a statistician and an architect - I accept that other people have greater experience and knowledge in certain areas than I do. If I want to charge my lawyer with malpractice, it behooves me to check with another lawyer to see if my complaint has merit, or if I'm even articulating the complaint correctly before I show up in a courtroom.--[[User:Brossa|Brossa]] 10:39, 28 July 2008 (EDT)

::Brossa has made several excellent points. I would add that there is a disconnect between your primary claim and the list. You should rethink your thesis statement or choose different "flaws" that fit it better, since your primary claim that the flaws "negate [the paper's] claim that E. Coli bacteria underwent an evolutionary beneficial mutation" does not follow from those listed. With your flaws, with the possible exception of #4 depending on interpretation (and even then, where did the undetected Cit+ cells come from?), you are arguing against historical contingency, not whether or not Cit+ cells evolved. In other words, the flaws you claim are questions of mechanism, i.e. the dynamics of how the Cit+ cells arose, but your thesis statement questions whether or not they arose at all from Cit- populations. So even if those flaws were true, they would not negate the "evolutionary beneficial mutation" as you say but rather just historical contingency as a path to it. [[User:Kallium|Kallium]] 21:23, 27 July 2008 (EDT)

::Brossa has made several excellent points. I would add that there is a disconnect between your primary claim and the list. You should rethink your thesis statement or choose different "flaws" that fit it better, since your primary claim that the flaws "negate [the paper's] claim that E. Coli bacteria underwent an evolutionary beneficial mutation" does not follow from those listed. With your flaws, with the possible exception of #4 depending on interpretation (and even then, where did the undetected Cit+ cells come from?), you are arguing against historical contingency, not whether or not Cit+ cells evolved. In other words, the flaws you claim are questions of mechanism, i.e. the dynamics of how the Cit+ cells arose, but your thesis statement questions whether or not they arose at all from Cit- populations. So even if those flaws were true, they would not negate the "evolutionary beneficial mutation" as you say but rather just historical contingency as a path to it. [[User:Kallium|Kallium]] 21:23, 27 July 2008 (EDT)