News

The following is a transcription of Dr. Gastonguay's acceptance speech when presented with the 2018 Bristol-Myers Squibb Mentorship in Clinical Pharmacology Award at the American College of Clinical Pharmacology on September 24, 2018.

Open Science: A Key to the Future Growth of Clinical Pharmacology.

It is a humbling honor to receive this award, considering the many excellent mentors in our field. To receive an award for mentorship is particularly meaningful to me.

 Thank you to the awards committee, those who assembled my nomination, the conference organizers, and BMS for sponsoring the award.

As I was preparing for this talk, I came across a view on mentorship that resonated well with me.

Ever Garrison said:

 "A teacher is a compass that activates the magnets of curiosity, knowledge, and wisdom in the pupils."

 I’m not so certain the physical science analogy alluded to in this statement is entirely accurate, but the idea that teachers activate curiosity, knowledge, and wisdom in their students seems to be right on target.

 I experienced this sentiment to be true through the example of my own mentors. Sorell Schwartz, my doctoral thesis advisor, taught me much of what I know about research, philosophy, and the blend of career and family life, and Tom Ludden, my postdoctoral advisor was particularly adept at creating opportunities to activate curiosity, knowledge, and wisdom for his trainees. There are many other mentors who impacted my career and my own mentorship habits - unfortunately too many to mention here. I am grateful to all of them.

 Thank you to all of the students, fellows, academic faculty, industry and government scientists I’ve been privileged to collaborate with over the past 25 years or so. Many of you are here today. I’ve learned from all of you.

 I acknowledge my amazing colleagues at Metrum Research Group. Whatever professional accomplishment I’ve been associated with in the last 14 years has been a result of our team effort.

And most importantly, I thank my wife Maureen, and children, Maddie and Chris for their tireless support.

 In the time I have left at the podium, I’d like to share some thoughts on a topic that matters deeply to me, and possibly ignite a call to action.

 What do I mean by Open Science?

The following key characteristics apply. Science that is:

• Accessible - to scientists and consumers at all levels, institutions, and regions
• Transparent - with respect to objectives, motivation, potential biases
• Reproducible - technical transparency about methods and assumptions
• Collaborative - this naturally results from application of the other three characteristics in an engaged scientific community

...If you think about it, aren’t these the basic tenets of science in general?

 In my experience in the domain of modeling and simulation, opportunities for open science include:

• Open Courseware
• Open Data
• Open ToolsOpen Models
• Open Publications

There are a growing number of examples of each, but significant opportunities for improvement and real challenges still exist.

Open and accessible courseware is essential for continued growth of the science and mentorship of our future leaders. In a multidisciplinary field, it is nearly impossible for individual mentors to cover the necessary breadth of instruction. Efforts aimed at creating shared learning content will benefit the entire discipline. We’ve learned this first hand from our own experience with Metrum open courseware and videos - hundreds of hours of content engaging 1000’s of viewers globally. This includes reaching regions where it may otherwise be difficult to access this type of content.

Open access to clinical trial data… Clinical research has made significant progress in this area. Initiatives like OpenTrials, OpenFDA are a step in the right direction, but we still have more to do. This is particularly true in the pre-competitive space for unmet medical needs, such as rare diseases. Open and accessible individual-level data sets describing natural history of disease progression should be our goal. We also need to be vigilant to potential challenges imposed by inadvertently restrictive data privacy regulations. Strong voices are needed to advocate for the scientific value of patient-level data.

Open tools and models are also essential. These are opportunities to hit all four characteristics; science should be accessible, transparent, reproducible, and collaborative. Physiologically based quantitative systems pharmacology models are one example. Clear, transparent, and accessible presentation of assumptions and all model details allow for community-vetted science and more credible results. Whereas, closed, black box methods provide little opportunity to grow the science. Checking the box of validation of these sorts of closed systems based on concordance of predictions with a few observations almost absurdly implies that we have no more to learn. Open models foster collaboration and acknowledge that we always have more to learn. Open tools allow for the easy extension of open models across institutions, infrastructures, and applications.

And now, a call to action…

Our community has a responsibility to demand and support open science principles.

Continue to use open source tools. We need to ensure the professional quality of these efforts and open access to supporting documentation. Participate in the development process and contribute feedback to developers.

When using open models, assess the transparency of modeling details and assumptions. Openly engage in the public vetting process. Challenge the authors if documentation is not complete or inconsistent with your observations. Suggest modifications or point to new evidence to improve models.

When you participate as a peer reviewer of research publications, consider the transparency and reproducibility of the work described. Trial designs should be completely specified, laboratory methods should be detailed or referenced, mathematical descriptions should allow the complete recreation of models by others qualified in the domain. I understand concerns about protecting intellectual property and that’s what trade secrets and internal documentation are for. Publication of research findings should adhere to the reproducibility principle. Otherwise, the credibility of the work will be suspect and an opportunity to grow the science is lost.

Seek to engage in pre-competitive collaboration on open data initiatives that are relevant to your research. Speak up for science when faced with the challenges of data privacy regulation policies in your institution.

Together we can grow the science of clinical pharmacology through accessible, transparent, reproducible, and collaborative open science methods. I am confident that this will lead to better science and better treatments for patients.

Thank you.