The way science keeps returning to the straight and true is by noting errors and by correcting them. Michael Shermer wrote an interesting article in the October 2001 Scientific American (page 30) entitled "I Was Wrong" He says "Those three words often separate the scientific pros from the posers". One can always correct one's errors and change categories, so posers can turn into pros. This page was written in response to critiques of the Ev paper and program.

To drive the point home, it seems reasonable to list known errata and corrigenda for our own papers.

Errata and corrigenda are corrections to typographical errors that can either come from the original text or are introduced by typesetters when they retypeset a paper. When a significant error was not caught until after publication, it can be reported in a correngium.

Obviously a deeper error is to publish a hypothesis or theory and it is later shown to be wrong. Of course there is nothing bad about that, it's the way science works. But it is nice to make a correct hypothesis or theory once in a while, and someone who consistently fails to do so probably won't last long in science unless they make really fruitful bad hypotheses! Some people make hypotheses and then stick with them in the face of counter evidence. An example is the hypothesis of Behe, who proposed in his book "Darwin's Black Box" that there are biological structures too complex to be evolved. The ev paper demonstrates a case that fits his criteria exactly but is indeed evolved. So his thesis failed, though he has yet to admit it as of this writing (2002 June 6 to 2004 September 29). However, from this demolished idea comes the idea that there are Roman arch-like structures in biology. According to Behe they could not exist since removal of any one stone would destroy the whole. Yet they do exist and are actually quite common in biology. A really simple direct example is an animal walking around on legs. Remove any one joint and it could not walk! How did it come to the arch of its body above gravity? It is pretty clear that one slithers along the ground first! This is an example of a dead hypothesis that leads to some interesting biology anyway.

Schneider Lab List of Errors and Corrections:

• Erratum to an equation in the 1991 edmm paper.
• Erratum to Figure 2 legend in the 1997 walker paper.
• 2002 June 9. Erratum for the 1998 rfs paper.
• 2002 January 18. Erratum of a reference in the 2001 flexrbs paper.
• 2002 June 4. Corrigendum to a figure in the 2001 repan3 paper.
• 2003 August 28. Corrigendum to text in the 1996 malign paper.
  2003 Sep 1: John S. Garavelli reports that there is a similar error in Monod's Chance and Necessity. This is a bit of text from John's PhD thesis

  @phdthesis{Garavelli75,
  author = "John S. Garavelli",
  title = "Part I. Studies on the Structure of Mitochondrial Malate
  Dehydrogenase Part II. Applications of Information Theory to Protein
  Structure",
  year = 1975,
  school = "Washington University",
  address = "Saint Louis, MO",
  type = "{Ph.~D.} dissertation",
  note = "{\it Diss.\ Abstr.\ Int.\ B} {\bf 36} (1976), 4437-4438"}
  

  which points out the correction in this text and footnote:

  "The globular shape peculiar to a given protein --- the special shape required for its functional activity --- will therefore be in fact dictated by the sequence of residues in the fiber. However, and this is the important point, the quantity of information that would be needed to describe the entire three-dimensional structure of a protein is far greater than the amount of information defined by the sequence itself. For example, for a polypeptide 100 residues long the information (H) necessary to define the sequence would come to about 2000 (sic) bits (H=log₂ 20¹⁰⁰), whereas to define its three dimensional structure this sum of information would have to be supplemented by a great deal more, the exact amount being difficult to calculate."
  
  Footnote: (The sequence information relative to conformation is, as will be shown, in fact much less than log₂ 20¹⁰⁰ = 432.192 bits.) (Monod, 1971, pp. 93-94) Monod, J. (1971) Chance and Necessity, Alfred A. Knopf, Inc., New York.

  (Reproduced with permission.) Note: I would suggest not calling H the information, but rather, the uncertainty.
• In a paper critiquing the Ev program I.G.D. Strachan pointed out a bug in the ev code (Appendix B, PDF - BUT www.iscid.org no longer exists as of 2016 Mar 17). Michael Y. Levashov has confirmed the bug. (The bug will be corrected when I have confirmed it for myself.) See: Evaluation of `An Evaluation of "Ev"'.

Notes

2003 May 8: I found an interesting example in Nature (423 (1 May 2003) www.nature.com/nature p.21) in "Life's transistors" by Fred J. Sigworth:

Practically everyone in the ion-channel field (including myself) has imagined the S4 segment to be an alpha-helix ... But the results of MacKinnon and colleagues show that it is almost certainly wrong.

2005 May 30: Even the mainstream media admits errors. Newsweek dated today, we find A Letter to Our Readers which retracts a story they published.

... we got an important story wrong, and honor requires us to admit our mistake and redouble our efforts to make sure that nothing like this ever happens again. ... when we make a mistake---as institutions and individuals inevitably do---we will confront it, correct it quickly and learn from the experience.