The wandering western blot of Mario Saad (which I reported on previously) was spotted yet again, no less than three times, which makes it now 15. The duodecuplicated western blot is therefore now upgraded to a quindecuplicated one. Also, some more of its new replication-happy friends emerged. A suspicion creeps in that Saad and his Brazilian colleagues José Carvalheira, Cláudio De Souza and Lício Velloso only ever made a handful of western blots which were forced to stand in for all possible instances in their many publications in high-profile journals. Most interestingly, Saad’s State University of Campinas in São Paolo (Universidade Estadual de Campinas, UNICAMP), apparently saw this as a perfectly normal research method. Saad, and as far as we know also his 3 partners, never had to answer for their creative approach to science. This of course may change, since Saad already had to retract a number of papers (currently six, according to Retraction Watch). Several of his publications also received editorial expressions of concern, which could mean further retractions. All this may be gradually decreasing Saad’s standing and funding-pulling value at UNICAMP.
The scientist whose recent analysis of Saad blot breeding I present here exclusively, is Paul S. Brookes, his is the entire credit for the figures below. His entire analysis is available here as PPT file.
Brookes studies heart mitochondria at the University of Rochester, USA, and is well known as a research integrity activist. He exposed many data manipulations on his website science-fraud.org, but was forced to close shop after his identity was revealed and certain questionable scientists set their lawyers upon him. Brookes has been investigating the data manipulation concerns in Saad’s papers as well. In March 2016, he demanded in an open letter that PLOS Biology finally retracts a Saad publication (Caricilli et al, 2001), “due to overwhelming evidence of data manipulation that has been known to you for almost 3 years” as well as excessive plagiarism. On May 23rd 2016, PLOS has finally retracted the paper for excessive data manipulation, despite that
“An institutional inquiry has been undertaken at the University of Campinas (UNICAMP) in São Paulo, Brazil and while the investigation acknowledged these errors, they did not find evidence of research misconduct”.
The images below list the 15 incarnations of the wandering western blot and align them to reveal the similarities. As Brookes put it:
“I think this makes a much more convincing case that these images all share a common ancestry. Notes in the 3rd slide are “idiot proof” to avoid claims by any editor that the evidence is not good enough”.
In addition, Brookes found further blot duplications:
“I looked more closely into one of the papers (Cintra et al, Life Sciences 2012 -LS), and tried to trace the heritage of some of the other images in that paper. […] There are 3 other blots in that paper that can be traced to others in the set of 10 papers. Again, some of these appear numerous times across the set (in a couple of cases the images appear in 4 separate papers, so this is not just coincidence). The one on the 4th page is pretty obvious, as is page 6. Not so sure about the one on page 5”.
The Cintra et al paper was already corrected on May 1st 2016, with the offending quindecuplicated western replaced by a different one. The publisher Elsevier may however need to fix other blot duplications in this paper soon, given Brookes’ fresh evidence.
Another two western blots were from this paper somehow stumbled into 5 other publications by the Saad team. One such cloned western Brookes found lurking also in the papers De Souza et al, FASEB J, 2007, Ropelle et al, PLOS One 2008 and Ropelle et al, Diabetes, 2008 (flagged in April 2016 with expression of concern due to other blot duplications). This western is therefore quadruplicated, appearing in four different instances.
Same goes for Ropelle et al, Endocrinology 2007 and Calisto et al, Critical Care 2012. The newly discovered wandering western blot is quadruplicated there as well, present twice in the former, once in the latter and once in Cintra et al, Life Sciences 2012.
There are of course other duplicated Saad westerns haunting the literature, for example this one, flagged by a pseudonymous commenter on my site. With so many blots mirrored across Saad et al papers, one should wonder if any of their research is any reliable. At some point journals might consider blacklisting these authors and issuing blanket expressions of concern or even retractions to their publications.
Mario Saad published in June 2019 on his university’s website a blog post, titled:
A clinical trial of fake news, by Mario Saad
“Refletindo sobre as intempéries pelas quais passei nos últimos quatro anos, me pergunto como foi possível que nós, pesquisadores brasileiros, sobrevivêssemos ao nosso trabalho, até o momento, sem uma estrutura desse porte no âmbito das universidades, considerando a realidade do desenvolvimento científico nacional?
Como numa espécie de ensaio clínico, sendo eu próprio, a cobaia, compartilho agora com os meus pares de academia, o drama de precisar provar que a integridade de seu trabalho é a razão de ser de seu próprio trabalho. Uma tarefa árdua, penosa, por vezes exaustiva e desanimadora, mas que com o empenho e dedicação, sobretudo daqueles que estão ao seu lado na bancada, pode, sim, ter um final feliz.
Introdução: Em meados de 2015, as revistas Diabetes, Plos One, Plos Biology, Critical Care e Diabetologia receberam denúncias anônimas apontando possíveis erros em figuras de meus manuscritos, sugerindo manipulação de imagens.
Objetivo: Demonstrar que as denúncias apresentadas às revistas científicas eram infundadas e que eventuais erros involuntários não podem ser confundidos com manipulação de imagens ou fraude em pesquisa.