Welcome: Julio Saez-Rodriguez | EMBL
/ via embl/
EMBL-EBI’s new Head of Research aims to rebuild the community post pandemic and leverage new AI technologies
Julio Saez-Rodriguez and his group develop and apply methods to analyse large omics datasets to understand what happens when disease sets in. As EMBL-EBI’s new Head of Research, he is also responsible for setting the vision, strategy, and scientific direction of the cutting-edge research taking place at the institute.
Shortly after he started in post, we caught up with him regarding his work on multiomics data analysis, his plans for creating a vibrant research community in a world of hybrid working, and his thoughts on leveraging AI to develop methods for solving global challenges.
What does your group’s research focus on?
My group extracts disease mechanisms from multiomics data, in particular single-cell and spatially resolved data. To do this, we combine machine learning, mathematical modelling, and prior scientific knowledge. The aim is to understand how cellular processes work at a molecular level and how these processes go wrong, for example, in cancer, cardiovascular, or autoimmune diseases.
We’re a computational group made up of biologists, physicists, computer scientists, and engineers, and we develop open source tools, which can also be used by other groups to explore different scientific questions. There is a big focus on open science and reproducibility.
What is the research environment at EMBL-EBI like?
I used to be a Research Group Leader at EMBL-EBI, and in my experience, EMBL is a special place for science because it’s young, collaborative, and interdisciplinary by design. For data scientists, EMBL-EBI is particularly unique because research sits very close to the public data resources managed by the institute, which provide high-quality data and curated knowledge to the world. These world-leading data resources are key for the development of AI algorithms such as AlphaFold.
Physically sharing a space with the teams managing these data resources is unique and mutually beneficial. It enables researchers to understand the complexities and context of data, while also creating a valuable feedback loop for the data resources. Ultimately, this helps to develop and improve the tools used in the services and the data shared with the community.
Crucially, because EMBL-EBI is a relatively small institute where most teams know each other, our researchers also work very closely with the admin and technology teams. The latter two play an essential role in supporting the robust and dynamic research taking place and allowing us to adapt quickly, which is vital in the fast-paced world of bioinformatics. In particular, EMBL-EBI’s top-notch and agile IT infrastructure is key to developing AI models.
What is your vision for research at EMBL-EBI?
I would like us to capitalise on our strengths and rebuild the community which has suffered during the pandemic and the cost of living crisis. I think it’s important to rekindle the energy and revitalise this vibrant place for doing science with lots of opportunities to talk to colleagues about your work. For me, a healthy research culture is also based on integrity and reproducibility, as well as being respectful to colleagues and creating a supportive environment that stimulates an exchange of ideas and interdisciplinary collaboration.
We’ll be recruiting two new group leaders in the coming months, which we hope will contribute to reinvigorating the research culture at EMBL-EBI. Equality, diversity, and inclusion are an absolute priority and we’re supportive of candidates regardless of their personal circumstances or background. In line with our research assessment efforts, we don’t need candidates to have a shiny big paper, but we’re looking for early-career researchers with an inspiring vision, good ideas, and a new angle on solving important problems in molecular biology. We’re particularly keen to bring on board people with a strong methodological emphasis, who can develop methods that can be used across a wide range of applications.
I think EMBL-EBI research will play a pivotal role within EMBL in the application of AI in biology. We have people at the cutting-edge of research working together with the data resources, groups in other EMBL sites (experimental as well as AI-focused groups) and transversal themes, as well as the wider scientific community locally in Cambridge and the UK, but also internationally.
What are some of the first things you’re hoping to do in your new role?
A big priority is to create opportunities for interactions between colleagues. I think this is at the core of successful interdisciplinary research. To some degree, we have to reignite the value of physically being on campus and chance encounters that simply don’t exist when working remotely. I think a hybrid working environment balances interactions and knowledge exchange with deep individual work. I’m very open to ideas and would love to hear from our colleagues and fellows what changes they would like to see.
How can EMBL-EBI address some of the sustainability and accessibility challenges of bioinformatics in the age of AI?
There are many opportunities arising from high-throughput technologies generating mountains of data and AI systems using these data to gain new insights. But there are also challenges. For example, AI systems require a lot of computing power, during both training and use. This means that the work has a significant carbon footprint, because all that computing uses energy. In addition, it creates barriers for people who don’t have access to high-end computing, which in itself creates higher inequality.
Accessibility has always been a priority for EMBL-EBI’s data resources, which are designed to serve the global research community. The same should apply to research whenever possible. We can do this by developing data analysis methods that can get similar results while using less computing resources. Going forward, developing efficient algorithms is, in my opinion, an exciting and important research area.
Why is EMBL-EBI’s research important to the broader society?
This is a key question. We don’t develop algorithms for their own sake. We develop methods that help us understand the biological processes happening in our world, to explain how life works. Only by doing this, we can start to develop solutions for the big challenges. We can improve our understanding of health and disease and ultimately create better patient diagnosis, treatment, and monitoring tools to support precision medicine. We can also understand biodiversity and climate change better and help to develop innovative solutions for key environmental challenges such as plastic degradation and upcycling waste from industrial processes.
The possibilities of bioinformatics and AI in biology are endless and it’s our role to explore them both in the lab and in conversation with society more widely. Engaging with the public is an area EMBL has already done a lot of work in, often focusing on underrepresented communities that don’t have regular access to science.
Maintaining an open dialogue with the public is essential if we want the benefits of bioinformatics to be universally distributed and well-understood, and this is vital in a world where misinformation is increasingly common.