Tackling the reproducibility crisis using blockchain technology

Scientists often find themselves in a situation where they are unable to reproduce the work of other researchers. Following one false lead after another can be very costly for graduate students or postdoctoral researchers looking to complete their research projects and advance their careers. As a postdoctoral fellow at Harvard Medical School, I spent more than one year breeding a mouse strain in which you could conditionally induce pancreatic cancer with a 100% penetrance. Or so I had deduced from the publication where this strain was first described.

I later found that just about 35% of these mice developed tumors. Because my project required that the majority of the animals acquired tumors equally, I had to look for another mouse strain and start over.

While the reported data in the paper was not false, it was misleading since the phenotype was dependent on the mice’s genetic background. I only learned this after discussing the problem with other scientists who had experienced the same situation in their labs, including the author of the reference paper. This data was not made public anywhere, and had this information been published, it would have saved me and many other researchers a great deal of time and money wasted breeding this “defective” strain.

So why wasn’t this crucial information available for all interested scientists to see? One reason is the current publishing system that is driven by the inflated commercial interest of publishers who don’t deem this information as novel or “newsworthy.”

ScienceMatters, an open access digital publishing platform, is looking to change and revolutionize the publishing world by creating a one-stop-shop data archiving and publishing platform powered by blockchain technology. Using their EUREKA platform, researchers can publish all their positive, negative, as well as replication data in real-time for crowdsourced, expert peer review. This gives authors immediate authorship rights, but more importantly, it eliminates the need for their observations to fit into a well polished story. EUREKA provides equal incentives for researchers to publish confirmatory, contradictory, or negative observations by rewarding them with EUREKA tokens. As a result, this data is made available for the entire scientific community to see with the click of a button. This allows researchers to make better informed decisions about lines of research worth pursuing and lines that may lead to dead ends.

This is also of particular interest to pharmaceutical companies who spend a great deal of money on research and development. In the US alone, pharmaceutical companies waste $28 billion annually on irreproducible life sciences research. ScienceMatters will be the first to publish replication studies on the blockchain and use a patented algorithm to link the replications to the original work. This will make it easier for drug developers to confirm whether a certain drug molecule is worth testing in clinical trials.

The non sharing of data, like in the example above, is one aspect of how the current publishing system is failing the scientific community and contributing to the reproducibility crisis. Other factors, however, also play a major role in this crisis:

The pressure to fit data that sometimes doesn’t fit into a developing storyline drives researchers to omit ‘inconvenient truths’ and, in extreme cases, to manipulate the data. Publishing on the EUREKA platform ensures that archived data is immutable, version controlled, and securely distributed thus providing an ideal system for research data management that does not rely on the trust of either publishers or authors.

The peer review process forms the basis for research quality control by allowing scientists in the same field to provide feedback on how to improve the work. These reviewers also decide whether the work is deemed worthy of publication. Relying on two, at most three, reviewers to evaluate a body of work can be problematic as it opens the peer review system to potential personal, professional, and social biases, both intentional and unintentional. By allowing crowdsourced peer reviewing on the EUREKA platform we ensure that the review process is bias-free and based on merit alone.

The ability to reproduce results is one of the cornerstones of high quality research. Blockchain technology provides the ideal tools to fight the reproducibility crisis by permitting the instantaneous and open sharing of data, minimizing bias and making research findings tamper proof.