HUM-MOLGEN archive: ETHI, SPEC: Fraud at NIH/HGP

My heart goes out to Dr. Francis Collins of NIH and the graduate student now accused of faking lab results that were the basis of several papers published in respected scientific journals. God only knows what pressures to perform are brought to bear on the young graduate student of today. However, some aspects of the case, as far as I can gather the facts from a few newspaper articles, do suggest very positive steps that can and should be taken to improve the situation in the future. [1] Truth in authorship. Ever since I was a lad, decades ago, the head of the lab put his name at the top of the authors list of every paper coming out of that lab, whether he personally directed and supervised the work or knew little about it. The prestigious name at the top of the authors list guaranteed a good reception from peer reviews and publication by the best journals. If the research turns out to be brilliant, the name at the top of the paper basked in unearned glory. I can suggest three categories for authors whose names appear on scientific papers -- [HFI] -- haven't the foggiest idea about the research on which the paper is based, but I certainly want the credit if it turns out to be great research resulting in scientific leaps forward. [PER] -- perfunctory direction and supervision. Takes 20 minutes a week to talk with the researcher, or scan a couple of pages of the researcher's lab journal. Has a foggy idea of what the research is all about, but few of the details of the experiment or the methods used. [PRT] -- a real partner in the research, even though spends as little as 3-4 hours per week on it. Knows methods being used, design of experiment, watches lab work for an hour or so each week. [GEM] -- same as [PRT] above, but arranges for another researcher in the lab to duplicate experiments being conducted by first researcher. The researcher duplicating the experiment knows it is only 7-10 days of his work and that someone else will do it in the future for his experiment. Such duplication to be done every 5-6 months. The extra costs of doing this, when all the equipment and materials are available, seem small in comparison to the costs of e.g. the articles and furor now rising from the Francis/NIH situation. And such duplication in the lab would do much to keep essentially honest researchers on the straight and narrow path. ================= [2] Truth in peer review. Again, I suggest several categories of reviewer, to be added after the reviewer's signature, and as the basis for a footnote by the journal editors.-- [NON] -- name only. I signed the review, but really didn't read most of the paper, gave no thought to it. Since I also wasn't paid for this work, it's a fair bargain. [PER] -- perfunctory review. Actually read most of it, gave it some thoughts about how it relates to reviewer's life work. Writes a couple of paragraphs suggesting ways future experiments might be done. [BST] -- the researcher's nightmare. Read review carefully, thought about what could go astray on this kind of research, requested copies of the lab records for one week, selected at random, and suggests how happy the reviewer will be when competitive work in this area is done by another lab. The fact is, if one reviewer in the Francis/NIH case caught the error which led to the exposure from the paper itself, the possibility is raised that the other reviewers, if they were doing their job, should have caught it also. But they didn't. The publication editor could then add a note to the title of each paper published, indicating whether peer review was seriously carried out, or not. The above suggestions, while tinged with a little humor, are serious suggestions. They could be implemented without a lot of problems or preparations, with relatively small cost, and would immediately help alleviate the occurence of scientific fraud in the lab. Gene O'Regon <gene@sqn.com>