The State of Open Science, and What’s Ahead
Six Mozilla Fellows reflect on their work in the realm of open science
This post is by Mozilla Fellows Kadija Ferryman, Julia Lowndes, André Maia Chagas, Ciera Martinez, Alexander Morley, and Daniela Saderi
We are coders, researchers, and community builders united by the vision of better, more open science. Our backgrounds as researchers have prepared us to tackle openness in software, hardware, ethics, data, skills, and publishing. We know it is not enough to be performing science in an open way; we must also encourage others to do the same.
In February 2019 we met for a week together in New York City, hosted by the Siegel Family Endowment and the Helmsley Charitable Trust. Halfway through our Mozilla Fellowship, we had several goals for the week. We came together to strengthen relationships with each other, discuss our projects, identify ways we can collaborate and support each other, and create presentations for an evening of lightning talks. We also built a website to literally put all of our open science work on the same page: mozillasciencefellows.com. In this post we will summarize what we discussed in those presentations.
Open software
I’m Alex, as a 16 year old I took antidepressants for depression, and I still can’t believe the effect they had. But they don’t work for everyone, and as a neuroscientist my goal was to make some progress in figuring out how they worked so we could develop better treatments. But the deeper I dived, the more I felt like everyone was running on a treadmill, working hard but working in a system that wasn’t built for moving forward. The problem is multi-faceted, but as a group of fellows we are coming at the issue with a diverse set of foci. Mine is raising awareness of software as not only a critical tool for research, but also a valid output in and of itself. Without this the building blocks on which we base our conclusions are unstable and incremental progress is very hard to come by.
Recently, as part of this effort, I’ve been focussing on the sustainability of open scientific software, in particular, how we can use estimates of the financial value of an open project in order to garner support from key stakeholders (see Back Your Scientific Stack). I’ve also been looking at how human-centered design principles can be applied to research software in order to bolster its usability and inclusivity; this is being put into practice in the Graphical Notebooks project.
Part of the original vision for this fellowship was to work on tools that enable automatic integration of software, data and visualisations which we referred to as “Continuous Research”. This work has been carried out in collaboration with the Flatiron Institute in New York, where we have built a pipeline for automatically benchmarking new software and datasets and publishing the results to the web (see Spikeforest). Some of the learnings from this project will also likely make it into The Turing Way: a how to guide for reproducible data science which is a book being written at The Alan Turing Institute for data science in London.
Open hardware
Hi I’m Andre, a neuroscientist interested in increasing access to science and education tools. This interest started when I was conducting experiments that required equipment valued over 60k Euros. At the time, I was studying in Germany and had plans to go back to South America. Two things struck me at the time: a) a lot of the differences in research output between different countries are related to infrastructure problems (and not differences in individual capabilities), and b) the hardest issues to circumvent were related to hardware access (laboratory tools, reagents, computers), as there was no way of asking someone for a pdf copy of a centrifuge, or using a pirate version of a microscope.
Luckily technology has evolved to the point where very powerful electronic building blocks have reached the consumer market and internet infrastructure allows us to share tutorials and blue prints in repositories, video platforms and online forums, boosting the maker movement.
My project with Mozilla aims to push open science by building low-cost but high-performance scientific equipment based on Open Source Hardware. Together with different groups and labs we are building a global understanding of current research-tool needs that are not met by competitive open hardware alternatives to kick-start their production together with makers across the globe.
Open ethics
I’m Kadija, I’m an anthropologist. I conduct research on the ethics and social impacts of precision medicine. My work provides a critical perspective on the development of health technology, especially focused on issues of inclusion, power, and social justice in data-driven biomedical research. This interest was sparked by a comment I heard a physician make when I was a teenager. He said that women from certain countries have smaller veins than others. This set off a cascade of questions that push me to this day — how does race matter in health? How is medical practice informed by research? How do our social and cultural values become part of health care? What does it mean to do “good” in medicine and medical research?
My research aims to address these questions, and my most recent work addresses the emerging field of data-driven precision medicine. This field of biomedical research and medicine aims to use multiple forms of data, from genomics to electronic health records to deliver personalized healthcare. This sounds like a laudable goal, but as I conducted more research in this area, I realized that precision medicine could have downsides, especially for marginalized groups, and there was limited research exploring the potential negative impacts.
I launched the Fairness in Precision Medicine study to examine how health data can be biased upstream, and how there can be cascading effects of bias that can lead to downstream discriminatory outcomes. This work has received attention in the popular press and has expanded the conversation on precision medicine to include critical perspectives on the potential for bias.
As a Mozilla Fellow, I am focusing on how we can make precision medicine more inclusive, how we can make embedded biases more visible, and how we can avoid pitfalls. I recently published an editorial describing how artificial intelligence in health can exacerbate health disparities and the steps we can take to mitigate that potential, as well as delivered a keynote address at the Open Con Cascadia conference on how precision medicine can use principles from open science to achieve more just outcomes. I am also working on a paper that discusses inclusion and diversity in precision medicine research, as well as case studies that explore how we can identify and avoid data bias and discriminatory outcomes.
We are at the precipice of precision medicine — it is still early enough to shape the development and direction of the field. It’s my goal to work towards a future of health that embraces openness, inclusion, and equity.
Open data
I’m Ciera, an Evolutionary Biologist obsessed with using large amounts of data to understand how life evolves on this planet. One of the most exciting aspects about being a researcher at this moment in time is the transparency we are now able to achieve in our data and research practices, allowing a level of reproducibility previously not possible in science. The practices and technologies now available for handling and sharing data, allow researchers for the first time to truly build off each other work, thus enabling better and faster science. I have made it my mission to help build upon these technologies and practices in both how I practice my research and through my open data advocacy work. I have helped establishing best practices, data standards, and in the creation of education materials all in an aim to make the scientific community more inclusive and efficient.
This year I have been excited by one of humanity’s most precious resources — data from Natural History Museums and Botanical Gardens. These collections have been accumulating for hundreds of years offering billions of specimens for researchers to use for research describing our ever changing planet and revealing patterns of evolutionary history. I have started the Cabinet of Curiosity project in an aim to make natural history and biodiversity data more usable, accessible, and valued. Visit curiositydata.org for tutorials, tools, interviews, and collection visits. Do you work on biodiversity data and education? Come join our efforts! You can follow our upcoming work and contribute through our Github organization and on Twitter.
Open skills
I’m Julie, a marine ecologist. I work to build and strengthen communities of practice around environmental open data science. Openscapes is a program I just launched as a Mozilla Fellow to engage, empower, and amplify scientists. It centers around a remote mentorship “Champions” program for labs with a lesson series building from Lowndes et al. 2017.
The Champions program is meant to help these teams become more efficient on the data analysis side of things, so they can uncover environmental solutions faster. Our inaugural cohort is really exciting. It includes seven environmental scientist Champions and their labs, totalling 24 scientists. The group includes marine and freshwater scientists, faculty and lecturers, postdocs, grad students, lab managers, and technicians. They are studying critical and time-sensitive topics, from climate changes’ effects on global food systems to parasitology, toxicology, and human health. The program is led remotely over five months, and is modeled after Mozilla Open Leaders. Each month in addition to one-on-one mentorship with each Champion, we all meet together twice to learn tools and discuss how to reduce friction when doing collaborative data analysis in the lab — and in science broadly.
Openscapes.org is meant to be a friendly landing page for anyone who might be interested in open data science but not sure how to engage. All our blogs, resources, presentations, and lesson series are meant to help, with more coming! And we are active on Twitter!
Open publishing
I’m Daniela, a neuroscientist. I live in Portland, Oregon, but I was born and raised in Sardinia, a beautiful island in the heart of the Mediterranean Sea. I am an open science advocate, passionate about research and making scientific output available to everyone to advance knowledge and improve transparency. I recently co-founded PREreview, a community and an open platform to facilitate the collaborative writing of preprint reviews and the training of early-career researchers in scientific peer review.
Preprints are early, yet complete, versions of scientific manuscripts made available to the community before editorial-organized peer review. As open access scientific manuscripts, they represent an invaluable wealth of data and knowledge made available when community feedback is most useful. Despite the many positive aspects of posting preprints, including speeding up knowledge dissemination and discovery, many researchers in the life sciences have not yet widely accepted them as a valuable way of sharing scientific output.
With our work at PREreview, we want help drive a cultural shift in which every scientist posts, reads, and engages with preprints as standard practice in scholarly publishing. In particular, I want to help scientists see the benefits of sharing openly their work and exchanging feedback, in a way that is constructive, inclusive, and rewarding to them.
This project is led by Samantha Hindle, Monica Granados, and myself, and it is fiscally sponsored by the non-profit organization Code for Science and Society. We recently received funding from the Alfred P. Sloan Foundation to build on a new, interoperable open source PREreview platform, designed from the ground up with our community in mind. We were also the recipients of the the Wellcome Trust’s Open Research Fund to develop Rapid PREreview, a tool that will allow scientists to swiftly share feedback to preprints during public health crises, and also to generate aggregated data visualizations based on feedback.
