Edit

We are EMBL: Francesco Costa on protein design and the EMBL PhD programme

Francesco Costa talks about his work on protein structure predictions, his experience with the EMBL PhD programme, and why he's excited about the field of protein engineering

Man in green jacket leaning against a grey wall.
Francesco Costa is currently doing his PhD at EMBL-EBI. Credit: Jeff Dowling/EMBL-EBI

Francesco Costa is a PhD student at EMBL-EBI, focusing on protein science. He is fascinated by protein design and has been working to improve protein structure predictions and explore proteins that might be interesting for industrial applications. 

Tell me a little bit about your background.

I’m originally from Italy and studied biotechnology at the University of Padua. I did two Erasmus internships, one at the Technical University of Munich, and one at the University of Bonn. I then applied for EMBL’s international PhD programme and in 2022, I started at EMBL-EBI, in the Bateman group

What does your research focus on?

I’ve got two main research directions: one is about improving protein structure models, such as the ones created using AlphaFold, and the other is to study fibrillar adhesins, which are proteins used by bacteria to attach to surfaces and interact with other bacteria. 

I recently published a paper on ‘rescuing’ low confidence AlphaFold predictions, using protein family templates. Protein families are groups of proteins that have very similar structures. We found that the Nobel-prize-winning AlphaFold2 technology predicts members of the same family with variable levels of confidence, usually either very high or very low confidence. We used information from the InterPro database run by EMBL-EBI to improve low-confidence models. We found that this approach improved the confidence of about one-third of the predictions we looked at. Being able to speak to colleagues who maintain the Pfam, PDBe, and AlphaFold databases helped me gain lots of insights on how to interpret the data and what to look for specifically. 

Moving on to fibrillar adhesins – my second research interest – these are very long proteins, made of many repeated copies of the same domain, called the stalk domain. Evidence suggests stalk domains have developed some kind of resistance to misfolding, meaning that they do not lose their 3D shape even after you heat the proteins in solution and unfold them. Lately, I’ve investigated the presence of covalent bonds that link specific residues and which help stabilise a protein’s structure. These isopeptide bonds are essential to protein structure stability and relevant for industrial applications. My colleagues and I developed a tool to predict the presence of these intramolecular isopeptide bonds and we’re keen to study them further. 

What are you most excited about in your field?

I’m really excited about all the things we can do with protein design and engineering. I think nowadays we have the tools to predict the structure – and maybe to some extent even the behaviour – of proteins, which is very exciting. 

For instance, researchers have created protein nanoparticles, which are similar to the protein shells that enclose a virus’ genetic material. This technology could be useful for developing new vaccines. 

What are the barriers to protein design and engineering?

We need more high-quality, reliable experimental data on specific design challenges. These days we’re seeing more and more blind competitions, similar to the CASP challenge which gave rise to AlphaFold. Researchers are asked to submit their solutions for a protein design problem. Then someone does all the experiments and provides a result. This gives us a way of determining which proposed method works best. 

What are some of the things you enjoy about life at EMBL?

Lots of things, for example, coming to campus by bike in the summer. It’s a lovely cycle path and the countryside is beautiful. 

Also, I enjoy the connection with other people who work in similar areas and learning by discussing your research project with them. EMBL has an incredible ability to bring together people working in the same field. This is incredibly valuable when you’re just starting out. 

What advice would you give someone who is thinking of applying to the EMBL International PhD programme?

Don’t get discouraged if you don’t have a strictly bioinformatics background. One of EMBL-EBI’s greatest strengths lies in the diversity of its people’s expertise, which helps foster creativity and innovation. At EMBL-EBI, you’ll find an open and collaborative environment that will help you acquire the skills necessary for pursuing your PhD. Whether it’s a bioinformatics hackathon or a coding course, there are plenty of opportunities to learn and improve your knowledge.

Do you have any unusual hobbies or interests?

I enjoy doing film photography – especially portraits – and I’m currently learning how to develop my own photos. 

EMBL's International PhD Programme

Applications are open until 10 March 2025 for EMBL’s International PhD programme – visit the link above to find out more. The programme provides comprehensive interdisciplinary training, maintaining a careful balance between theory and practice, close mentoring and creative freedom, collaborative teamwork and independence.


Tags: alphafold, bateman, bioinformatics, embl-ebi, interpro, open data, protein sequence, protein structure, we are embl

News archive

E-newsletter archive

EMBLetc archive

News archive

For press

Contact the Press Office
Edit