Building community
A conversation with co-affiliated group leaders at the VIB Center for AI & Computational Biology
September 11, 2024
The modern life sciences are driven by collaboration. As researchers unravel more and more of life’s mysteries, they increasingly depend on each other’s expertise to find answers. The advent of new technologies has the same networking effect, and machine learning is a prime example of this – AI has applications across levels of biological organization, from genes, over tissues, to organisms and populations. VIB’s Center for AI & Computational Biology (VIB.AI) uses the latest computational and AI strategies to address biological questions and challenges that span different disciplines. As a result, the center has built a community of VIB PIs from other VIB research centers, which pioneers a way of working that embodies science’s increasing emphasis on collaboration and interdisciplinarity. But what does such a co-affiliation mean for their work?
Let’s ask Anastassia Vorobieva (VIB-VUB Center for Structural Biology), Steven Maere (VIB-UGent Center for Plant Systems Biology), and Yvan Saeys (VIB-UGent Center for Inflammation research).
What does it mean to be a co-affiliated PI at VIB's new center, VIB.AI? Do you regularly meet with the other PIs, for example?
Anastassia: “As a co-affiliate PI, I take part in the group leader (GL) meetings of VIB.AI, regularly communicate through the Slack channel, and my students can present their work at the joint AI meetings attended by groups in all Flemish universities. I've had the opportunity to contribute to the GL recruitment of the VIB.AI center and I also try to contribute to community building by opening up the AI-related invited talks in Brussels to the rest of the VIB.AI community. Finally, my group has early access to VIB.AI infrastructures.”
Steven: “At the moment, this mainly means that my group participates in the VIB.AI Bioinformatics & AI seminar series, which is very useful to get to know more about what is happening at VIB.AI and the co-affiliated labs, and to get feedback on our own research from people with a machine learning rather than a molecular biology background.”
Yvan: “What it means to me is a lot of excitement; excitement to be part of setting up and designing a new VIB center from scratch and being able to choose our new colleagues. We have monthly faculty meetings with all the VIB.AI PIs and there is a great collaborative atmosphere and enthusiasm to work closely together in the future. It is really nice to have a new center dedicated to computational biology and with all the recent advances in AI we can be confident that the new center will become an important stronghold in VIB for all data science-related matters.”
.png "Anastassia Vorobieva, Steven Maere, and Yvan Saeys.")
Do you think that a co-affiliation reflects the interdisciplinary nature of your research? In your view, is this a larger trend in which new technologies can break down disciplinary walls?
Steven: “It sure does reflect the fact that plant science and other life sciences are becoming increasingly interdisciplinary. The founding of VIB.AI is a recognition of the importance of the computational biology and machine learning disciplines in the life sciences.”
Yvan: “The co-affiliation is a way to be part of the center and get connected to both existing and new PIs. It is the ideal way to bridge interactions between the new VIB.AI center and the existing VIB departments to which we are already affiliated. By design, computational biology is already an interdisciplinary field, so all PIs (co-affiliated or new) already have a mindset of breaking barriers between disciplines and going beyond dogmatic thinking.”
Anastassia: “I think it becomes difficult to do impactful research without interdisciplinarity and collaborations. In particular, in my field - computational protein design – it is critical to combine AI models and wet lab research to verify the computational models and inform the next design experiments.”
How do you hope to integrate machine learning into your (future) research?
Anastassia: “We are already using machine learning (and other physics-based) molecular models in our research. However, it is difficult to keep up with the fast-evolving field of computational protein design, especially since we are one of the few labs in the world that designs membrane proteins and we need to simultaneously build innovative computational pipelines and set up complex experiments just to be able to test our designs. That's a lot for a recently started lab. We hope that the co-affiliation to VIB.AI will help us to build more momentum and develop new ideas, which would be more challenging to achieve in isolation.”
Yvan: “We are a core machine learning team, so machine learning already prominently features in all our research lines. What is interesting is the fact that we are now also connected to other computational biology labs, some of them also heavily investing in more methodological machine learning research, and we look forward to connecting more with them, as well as with the more applied labs of VIB.AI.”
Steven: “Likewise, we’re already heavy users of machine learning and we also do quite some method development in this area, mostly on modifying existing algorithms to, for example, better take the spatial context of field trials into account or to provide confidence intervals for predictive model performance scores.”
Thanks, everyone!
As VIB.AI expands, the co-affiliated PIs will play a crucial role in building the center’s community and charting the course to further build a robust AI and computational biology network within VIB and beyond.
Want to be kept up-to-date on our biotechnological news and stories? Join our community and subscribe to our bi-monthly newsletter.