There’s no eureka in science

Published July 01 2017

In April, thousands gathered on the National Mall in Washington, D.C., to march for science, exhorting continued innovation and advocating for support despite growing skepticism.

Confidence has eroded due to flagging enthusiasm as public figures decry overwhelming evidence, using selected studies to repudiate the conclusions of the scientific community. This is exacerbated further by news media that often focus on breakthrough potential and fail to capture the uncertainty inherent in a single study. As a result, there is growing sentiment that biomedical research has failed to the deliver on the promise of new therapies. Many criticize scientists for an inability to translate initial findings into medicine.

It is therefore the responsibility of the scientific community to encourage direct outreach that emphasizes that there’s no eureka in science; discovery is a deliberate process toward consensus practiced cooperatively by an entire community of researchers.

Human disease pathology is incredibly complex, and while new therapies have been slow to arrive at the clinic, research has deepened our understanding of the fundamental biology that underlies pathology and dictates patient outcome.

More than 25 years ago, the National Institutes of Health began the Human Genome Project, a monumental collaborative effort to record the entire sequence of human DNA. The development of inexpensive high-throughput sequencing and computational biology techniques has allowed physicians and scientists to sequence the DNA of thousands more patients to identify mutations that drive disease pathology. This immense undertaking has changed fundamentally the way that scientists and physicians understand many diseases.

Physicians now can identify more accurately the underlying cause of disease and deliver targeted and personalized treatments for each patient. Furthermore, scientists have used these studies to identify novel targets for a whole new generation of drugs.

While progress has been slow and deliberate, the collaboration between patients, physicians and scientists ensures that every patient receives better care.

Despite recent advances, there are still real challenges facing scientists seeking to expedite drug development. One of the most challenging obstacles to the development of novel therapeutics is the interpretation of preclinical results to identify compounds for further drug development. Since the isolation of HeLa cells in the 1950s, human cells capable of growing in a dish have proved to be one of the most useful tools in understanding fundamental biological processes. Cell lines offer an inexpensive model of human biology that can be used quickly to test multiple hypotheses. However, human cell lines are imperfect models of complex biology and often cannot fully reflect disease pathology or predict patient response.

Molecular biology techniques developed in the 1980s allowed the genetic engineering of laboratory mice that more closely recapitulate disease pathology. More recently, the development of patient-derived xenografts demonstrated that scientists can interrogate patient tumors directly in a scientifically tractable manner.

Despite these efforts to improve laboratory models, only five in 5,000 preclinical drugs that show promise in the laboratory advance to clinical trials. Furthermore, four of those five either fail in clinical trials due to lack of safety or efficacy in human patients or are deferred for strategic reasons.

As the scientific community advocates for continued research, it is necessary to be honest about institutional shortcomings and introduce reform. With the growing prevalence of gene sequencing in biomedical research, a basic understanding of statistics and computational biology should be a required component of any graduate training curriculum. Additionally, as cost decreases, fingerprinting for every cell line used in a study should be a prerequisite for publication in a peer-reviewed journal. As scientific journals increasingly are published online, detailed protocols for novel techniques should be added to new papers in order to ensure reproducibility. Finally, peer review of papers and grants should be streamlined to ensure more efficient evaluation of new work.

Scientific discovery is a process of continuous evaluation, critique and evision. Research labs around the world with access to new and different cell lines or genetically engineered mice will use scientific literature to guide novel inquiry. After publication, the results are reviewed, challenged and revised by other researchers in the community. While this process is painstakingly slow, the scientific community eventually converges on consensus and discovers new biology.

There’s no eureka in science — there’s only progress.

Daniel Goulet Daniel Goulet is the Ph.D. candidate at the University of North Carolina, Chapel Hill, department of pharmacology.