A negative result is also a result

The crisis of reproducibility of oncological research. Act II

Julia Korowski, XX2 century

A major project to reproduce the results of scientific work in the field of oncology has released a second series of publications. Unlike the first series of experiments, this time the results are encouraging: not all the results were reproduced, but most of them. Articles-reports on new replication attempts are published in the journal eLife (Reproducibility Project: Cancer Biology).

In 2011 and 2012, scientists started talking about the fact that not all experiments in the field of oncology can be repeated. Specialists from the pharmaceutical company Bayer said that they were able to reproduce the results of only 25% of the selected preclinical studies, they were echoed by employees of Amgen – 89% are not reproduced. Scientists from the Center for Open Science and Science Exchange organizations did not like this state of affairs at all, and the following year they jointly undertook the repetition of other people's experiments. To begin with, 29 of the most important works on oncobiology published in 2010-2012 were selected, then 30 laboratories were hired. And in January of this year, the first bad news appeared.

The project participants managed to reproduce the results of only two out of five studies: in one case the results were negative, in the other two they could not be interpreted (there were problems with tumor models). Experts started talking about the "reproducibility crisis". However, new publications show that not everything is so bad – this time scientists were able to repeat most of the experiments described in two articles.

One of them said that mutations in the IDH1 and IDH2 genes, which develop in some types of leukemia, cause cells to produce a substance that stimulates tumor growth. Another is that BET inhibitors (a class of drugs) can stop the development of a certain type of leukemia. During the experiments, scientists confirmed that the compound I-BET151 kills human cells in vitro and suppresses the development of cancer in mice. However, rodents from the experimental group lived no longer than in the control group – in contrast to the original study.

Some experts note that the result does not refute the data on the effectiveness of I-BET151. They indicate that the final experiment was still different from the initial one, in particular, the animals received smaller doses of the drug. Oncobiologist and author of the original work Tony Kouzarides believes that the negative result "draws our attention to the pitfalls of biological research, namely, that conditions in different laboratories may differ, and this may affect the result of a particular experiment." Tim Errington from Open Science objects: if the experience "works" only under certain conditions, perhaps the result is not so reliable?

However, the eLife editor Sean Morrison calls the results of the project "encouraging." Concrete results were obtained in five cases, in four of them it was possible to reproduce the "essential aspects" of the original research.

Portal "Eternal youth" http://vechnayamolodost.ru  28.06.2017