sexta-feira, 17 de outubro de 2008

Fraude ou O QUE em estudo de células tronco ?

Oi pessoal
Continuando com nossa série de posts sobre fraude na ciência, vamos ao que deu na Nature dessa semana. Nem sempre o que parece fraude é, de fato, fabricação de resultados. As vezes é despreparo (ou pressa para publicar) puro e simples. O que realmente aconteceu neste caso ? Nunca saberemos de verdade.

'Manipulated' stem-cell paper faces retraction

The University of Minnesota has asked the journal Blood to retract a high-profile paper on adult stem cells following a university investigation. The paper reported that mesenchymal stem cells isolated from adult bone marrow could generate a surprising number of tissues (M. Reyes et al. Purification and ex vivo expansion of postnatal human marrow mesodermal progenitor cells. Blood 98, 2615–2625; 2001), but other labs had trouble replicating that work. The investigation concluded that the paper included manipulated images.
.
Lead author Catherine Verfaillie, now director of the Stem Cell Institute at the Catholic University of Leuven in Belgium, was cleared of academic misconduct but blamed for insufficient oversight. That suggests the blame rests with the only other scientist under investigation, Verfaillie's graduate student Morayma Reyes, now at the University of Washington. However, the findings regarding Reyes cannot be released because of privacy laws. Reyes says that she made "honest unintentional errors".
.
Problematic images were also identified in a patent and in articles published by the Journal of Clinical Investigation and Nature. But the university did not find sufficient evidence of misconduct in these incidents.
.
The University of Washington says it may decide to investigate Reyes.
.
Published online 15 October 2008
Nature vol. 455, 849 (2008).
.
===============================
Comentários dos leitores
.
Karthikeyan Narayanan
Pressure on researchers to publish and to get grants make biased judgemnts. However it is a huge mistake to fake and manipulate results. It is high time to reform the way we judge the published papers. One way of looking at it is the citation index. Citation index may not be the correct way to judge the reproducibility of anyone's work. But we should develop newer ways to judge a particular researcher's work by looking at new angles such as "reproducibility index" or "validity index". Both these indices would be solely on the follw-up research done by other labs or confirming results from other labs. This would (I think) help to judge some of the break through papers and their claims.
.
Miguel Martin:
It is absolutely true that impact factor is not all (even if all of us look for get it at least once in life, actually science is not about that...It is also true that the number of citations a paper has use to be associated with quality, impact, reproducibility...In biology, there are a number of classic journals, with a medium-low impact factor average, very classic ones, where everybody in the field, lot of times, for instance when you are a grad student looking for technical help and good methodology descriptions, you open those journals and get what you are looking for, that is, well described, highly reproducible methodology, well dictated results according with methodology...
.
We know impact factor publication is not all, still we pursue to get one in life. Science is much more than impact. Could you imagine Krebs, Cajal, Curie...worry about impact factor more than about following a passion for their work, a believe on pure science...? It would be fantastic if at international level, some sort of committee could review the ongoing work of a scientist, with certain type of forms or records in which anyone could describe the day to day of the work, the mistakes, the very good days, the discoveries...And so, get the chance to make official and public (saving private companies and/or works pursuing for patents exclusively) your efforts, your goals achieved and the many number of mistakes behind them...Yes, probably this is just a chimera, but still...Who never thought at least once something like that?
.
Francisco R. Villatoro:
Miguel, your proposal is exactly Open Notebooks based on wiki or blogs. See http://precedings.nature.com/documents/39/version/1 for more details.
.
B. More Critical:
After a quick look at the paper in question there are several problems with the figures that stand out (detailed below). 'Insufficient oversight' on the part of the PI is right... but where were the reviewers ?
.
Fig 3: two year donor lane pasted into the figure. Fig 6B: collagen western blot appears the same as Fig 5C actin western blot (center panel) once fipped horizontally. Fig.7B: both actin blots show an abnormal horizontal discontinuity. Fig 9C: lane MPC of the myosin western blot has been pasted into the figure. Fig 10C: lane MPC of the Tek western blot has been pasted into the figure. Fig 12B: signal of endothelial and ostoblast lanes enhanced relative to other lanes. Fig 12C: Southern blot endothelium lane brightened with respect to myoblast lane. Did I miss anything ?
.

2 comentários:

Anônimo disse...

Esse MHL é um idiota e hipócrita. Ele frauda resultados e posso provar.

aguardem,
Capitão MST

Felipe Rodrigues disse...

O último comentário postado, "B. More Critical", alerta para algo realmente grave. Marcelo, o grifo na última pergunta é muito bom. Mas tão bom quanto, seria um grifo na última pergunta do primeiro páragrafo: "Onde estavam os revisores?".

Somos constantemente ensinados a acreditar fielmente na Ciência, pelos seus critérios e sistematização. Infelizmente, algumas pessoas tentam burlar regras e fraudar resultados. Mas não seria a revisão por pares uma forma de evitar isso?! É uma pena descobrir que o conceito de networking (antigo QI - Quem Indica) também existe na Ciência.

Abraços.