Joana Pereira to join VIB.AI as new group leader
Exploring the protein universe
Pereira joins us from the Biozentrum at the University of Basel, where she leads the Protein Universe Atlas project as a senior postdoc and project leader in the group of Prof. Torsten Schwede. The Protein Universe Atlas aims to map a vast and largely uncharted domain of biological data by modeling the diversity and evolutionary relationships of natural proteins.
“My long-term research interests focus on the application of computer sciences, machine learning, and bioinformatics to explore the structure, diversity, and evolution of the ‘protein universe’,” says Pereira. “By applying complex network analysis, graph algorithms, and deep learning, I aim to develop approaches to better interpret the massive and ever-growing sequence, structure, and function data in public repositories.”
In doing so, Pereira wants to uncover unknown protein families and evolutionary mechanisms of biological significance. In a first, pilot implementation of the Protein Atlas approach, which used only about 10% of all cataloged natural proteins, Pereira and her collaborators were already able to identify at least 290 previously undescribed protein families.
About Joana Pereira
Originally from Portugal, Pereira holds a doctorate in Chemistry from the University of Hamburg and EMBL, with a focus on structural bioinformatics and macromolecular crystallography. She has seven years of postdoctoral experience at top institutions, including the Max Planck Institute for Developmental Biology and the Swiss Institute of Bioinformatics.
Hidden potential
At VIB.AI, Pereira wants to double down on her efforts of mapping the unknowns hidden in protein databases, especially those areas with potential biomedical and biotechnological relevance.
From functionally unknown protein families that are uniquely found in the human microbiome, to dark proteins that may affect virulence or defense systems—there is a lot of unexplored territory with potentially relevant proteins waiting to be uncovered.
“The ultimate goal is to cover the entire universe of natural proteins, and to predict, annotate, and classify those proteins which may have a direct or indirect impact on human health,” says Pereira. “With AlphaFold and ESMfold making hundreds of millions of protein structures readily available to the research community, we have massive datasets at our disposal to uncover new biological insights that could drive medical and biotechnological innovation.”
Building bridges
Stein Aerts, Scientific Director at VIB.AI, is looking forward to the start of the Pereira lab in February: “We are thrilled to have Joana join us at VIB.AI. Her expertise in AI-driven protein bioinformatics, combined with her collaborative spirit, will be instrumental in advancing our mission to explore the biological unknowns through artificial intelligence. I look forward to working closely with her and building exciting new bridges between VIB.AI and different experts in protein modeling and microbiology at VIB and KU Leuven.”
It’s precisely this embedded connection that VIB.AI envisions, which attracted Pereira to join the newly established research center. “We find ourselves in an era of unprecedented data generation,” she explains. “The overwhelming volume of data generated and the complexity of the methods being developed pose a challenge for non-AI-experts. There is a valuable opportunity for us at VIB.AI to efficiently address and close this gap, democratizing the use and interpretation of AI-based methods and their associated data across all life science areas.”
Vice versa, Pereira relies on specific expertise to complement her own as she navigates unchartered protein territory. “While my line of research is fundamentally computational, I do not envision my group without close collaborations with experimentalists. That’s one of the reasons I look forward to meeting the many new colleagues across VIB and KU Leuven in a few month’s time.”
Welcome, Joana!
_