Dr. Afshin Beheshti
NASA TOPS—Success Stories of Open Science Series
With a large network of people, you can do a lot of fun things in science: Q&A with Dr. Afshin Beheshti on open science practices
Dr. Afshin Beheshti is a bioinformatician and principal investigator at KBR at NASA's Ames Research Center. With a background in physics, he works on space biology projects related to microRNA (miRNA) and mitochondrial changes. However, he refuses to be a one-dimensional researcher, which led him to work on various topics including COVID-19/SARS-CoV-2, cancer, high altitude impact on biology, and traumatic brain injury. He leads the Multi-Omics Analysis Working Group (AWG) at NASA’s Gene Lab, where everyone collaborates to develop guidelines and ways to process and analyze Big Data. Through this model, his team discovered that mitochondrial dysfunction, when the mitochondria do not work as well as they should, is a key biological issue during spaceflight. In addition, he is the president of a nonprofit called COVID-19 International Research Team (COV-IRT).
What is your definition of open science?
Open science is open data and open community. First, open science is when data are free and made publicly available to the community and the public. For example, if I generate a dataset for a National Institute of Health (NIH)-funded project, it should become public because the taxpayers paid for this—it's not like a private company. Even if you can’t publish the results, you should still make it public after a period of time so that someone else will be able to come up with some useful analysis; you're wasting taxpayers’ money if you don’t share the data and use it to its full potential. And sharing is not just about the data; it could be tissues from animal experiments—anything to share with the public.
Open science is also the idea of the scientific community coming together—sharing ideas and working together. I think that's the main part of open science that people in biology don't do as much because they are afraid of backstabbing. A lot can be done when people come together and share everything, including grants. You're not taking away from someone else. I would consider open science as “the socialism of science.” Some people think socialism is a bad word, but it’s actually a good word. COV-IRT is a nonprofit with $0 in its bank account, but we all shared our resources, which include effort and resources that allowed us as a group to do amazing work.
So, open science is that whole grassroots idea of everyone coming together. You can't just have open data without people sharing or having an open community. If no one shares anything, you need to continually reinvent the wheel—it will be a waste of time. Without an open mindset, open data will go nowhere.
Why do you practice open science?
Actually, I've always done open science coming from physics. High energy physics involves billions of dollars, with hundreds of people working on experiments with synchrotrons or high energy colliders. Usually, you can find a list of 200 authors on a paper. There's no last author or the senior author—everyone knows what they're doing and gets credit for what they do. That's the whole physics mentality.
What steps are you taking to accelerate open science?
Since I work at KBR at NASA’s Ames Research Center remotely, I don't have a traditional lab situation. I set up a structure where I have a community of co-investigators and collaborators with the lab infrastructure. From my grants, we account for costs and resources to be shared and give credit to everyone who has worked on the projects. For my experiments, I typically have lots of biological tissues in my freezers that I originally did not intend to use for my grants, but I save them so I can share them with others with minimal red tape so that they don't have to pay for those experiments. Once they get the tissues, they work on their own dime, but we all get credit—whoever helped with those experiments. Such things expedite science. That's how collaborations are built. The whole “my lab” mentality should be tossed out—it’s outdated.
What challenges have you faced while practicing open science and open scholarship? What strategies did you use to overcome these challenges?
The biggest challenge is people with closed mindsets. Some people never change their minds. Just show by example. Once you start collaborating with people, you see how well it works. You realize that no one's backstabbing anyone or stealing data from anyone because everyone starts trusting each other. With my colleagues at COV-IRT, we're putting out a lot of nice papers together in high-impact journals. I have people who are not as active over a period of time but will start participating more after seeing our results. I always say yes to people participating—if you want to change your mind, you can come and play with us. I'm sure some people might take advantage of it. But I don’t cut them out completely. I just make it completely transparent to everyone involved. If you lose one person who doesn't want to play with the open science rules, that's fine—it’s their loss. That person just lost the entire community because everyone knew what was happening.
Funding is another hurdle, but that's not been the main issue because people usually share resources within the larger community. Eventually, we all apply for grants together and get funding together. The more you do that, the more funding opportunities come. But, a lot of governmental red tape slows down open science. Of course, you don't want to break any rules that are set for good reason. You still have to protect yourself because not everyone's always as open or honest, even though you do open science.
Have you made mistakes while practicing open science? How would you address them differently if you were to do it again?
For any kind of research, you're going to make mistakes. If you don't make mistakes, you don't learn. I think sometimes you can be too open. For example, if you pitch your idea to 200 people, they start chiming in. And, some people may misconstrue what you might have meant to do. Then, the project may seem to them as something you didn’t actually propose. Those people jump ship without actually discussing what the issues are and clarifying the misunderstanding for the final product. To avoid a situation like that, you want a subset of 10-15 people before passing important things by the bigger group. I think that's probably some of the biggest mistakes I made because I would be too eager to share my ideas.
Another thing is to note that some people might have nefarious reasons to take full credit for your data. Most people who you give tissues to are very grateful. They're excited, and we're excited. But some people take those ideas, especially if you don't publish them. I heard a story about a professor who published a paper using open data that was shared in a group years ago when they were a postdoc. But, in their paper, there's no reference to the person who originally shared the data. So, this professor had other goals to advance their own career and basically took advantage of the open science platform. Mistakes like that don't really hinder your open science, but it does make people more cautious about open science. People do that, unfortunately. The silver lining is that in situations like that, people in the open science community are aware of the truth.
So, I try to deploy Confidential Disclosure Agreements (CDA) or Non-Disclosure Agreements (NDA)to prevent all that. Also, publish early. Don't sit on it for 10 years. Your papers are your key to showing that you've done it first. When you submit a paper, put a preprint out, too. People even do that before they submit it to any journal, just so that the preprint is out there. Sometimes, the preprint never becomes a peer-reviewed paper. But it's still documented online to show people that you did it. Whether it was peer-reviewed or not doesn't matter.
How has open science improved your research? Are there other benefits you have experienced from practicing open science?
The more you're known, the more your network grows. The more you do open science, the more people want to collaborate with you. They know that's the guy that gives tissues away for free or has 50 people on their papers and supports students. It’s like that little light with the flies coming in. It's kind of like a beacon. With a large network of people, by default, you can do a lot of fun things in science. You get more publications. People promote different research ideas and build a nice comprehensive story together. All the pieces of the puzzle start to come together. You start learning the whole story that connects to everything. It's not just factor X. What about everything around that which you've ignored? When you're isolated, you never see that. Open science makes science more interesting.
What would you say to early career researchers who want to practice open science?
For students, leave the advisor if they are not doing open science. I know it's easier said than done—when you're a student, it's tough for you. So, the way to get around that is to talk to other labs and other graduate students in other departments as soon as you can. Talk to the people outside of that lab, and you can get the real stories.
For young faculty, don't pigeonhole to just one subject. Think broadly so that you can apply your ideas to many different fields. You will eventually find people with the same mindset. And join communities. If you want to get a lot of publications, you should collaborate. I've had one publication at least once a month, if not more than one. And there's more in the pipeline pending right now. Of course, you could be stuck in a place that doesn’t promote open science. Then, just leave that place—it should become so dry that they'll have to open up their rules and their old mindsets. There are plenty of people out there who have an open science mindset. But you have to find them. If there's a will, there's always a way. Again, joining big networks opens up opportunities. You could approach people and say you’re considering leaving the current institution. You could ask if people know about any positions in other institutions.
What are some of your favorite open science tools or resources that you’d like to share?
There’s a whole bunch of them. But, for citizen scientists, NASA’s Gene Lab provides visualization tools. Even someone who doesn't have any data skills can play around with raw data. The National Cancer Institute also has an interactive data portal where you can access the data from their Cancer Genome Atlas (TCGA) program even if you’re a non-scientist. Another cancer-related data website is cBioPortal. Cancer researchers have thought about how to make this friendly for people for many years. And the COVID Data Tracker at the Centers for Disease Control and Prevention (CDC) makes data easily accessible. GISAID is also a nice tool that a non-scientist could use for different kinds of viruses.
When it comes to tools, the bioinformatic community is all about open science because no one wants to pay for the proprietary software because they're expensive unless your university or your institution has a license to pay for them or you have grant money. These people are smart enough that they could create their own software and algorithms and then make them available. It might not be as good as the expensive software, but it's still good enough. And you could do a lot with it. Cytoscape is a free tool funded by an NIH project that you can download on your computer and plot networks. You can find examples on the homepage. It’s a nice tool because you don’t need to worry about programming. You can enter data and make neat networks out of it. There’s another tool called STRING, which you can use to look at protein interactions. These are just two examples of many.
Lastly, is there anything you would like to share regarding open science?
I think more people should do open science. Join my groups! I've been doing open science before it was a fad. I'm more than happy to collaborate with anyone from different creeds of life. As long as you are breathing and have a good idea, why not?