Welcome: Stephanie Lo
EMBL-EBI’s new Protein Function Content Team Leader brings extensive experience from the Wellcome Sanger Institute, where she led the Global Pneumococcal Sequencing project
Stephanie Lo has joined EMBL-EBI as the new Protein Function Content Team Leader. We spoke to Lo about her experience, her scientific journey into protein research, and how AI-assisted protein function annotation can accelerate biological discovery and support global health.
What is your background?
I previously worked at the Wellcome Sanger Institute, where I led the Global Pneumococcal Sequencing (GPS) project. This international initiative analysed genomic data from the bacterial pathogen Streptococcus pneumoniae, which is a major cause of pneumonia, to understand its evolution, spread, and antimicrobial resistance. This has helped inform vaccine design and surveillance strategies.
Growing up in Macao, I developed a strong interest in science and curiosity about understanding how the human body works. My scientific journey began with an undergraduate degree in Clinical Laboratory Science at Shanghai Jiao Tong University, China, which triggered my interest in microbiology and infectious diseases. I then pursued a PhD at the University of Hong Kong, where I studied antimicrobial resistance and the transmission of resistant bacteria between humans and animals.
Through my work in genomics, I became increasingly interested in proteins, because they are functional molecules that ultimately determine how cells behave.
Can you explain what the Protein Function Content team does, in simple terms?
Our team curates protein function information in UniProt, the world-leading source of freely accessible protein sequence and functional annotation data. In simple terms, we study proteins and describe what they do, how they work, and how they contribute to biological processes.
We use a structured framework called Gene Ontology, which provides standardised descriptions of protein functions, biological processes, and cellular locations.
The high-quality manual data curation that our team does is essential because it creates reliable reference datasets. These datasets are increasingly used to train machine learning and AI models, enabling automated annotation of millions of proteins. While a curator might manually annotate a limited number of proteins each week, AI-assisted approaches can scale this process while maintaining high standards of accuracy and consistency. Manual and automated data curation complement each other very well.
As you mentioned earlier, you have led the Global Pneumococcal Sequencing (GPS) project. How has this experience shaped your approach to protein research?
The GPS project focused on genomics, analysing DNA sequences to understand bacterial diversity and evolution. However, genes ultimately carry out their effects through proteins, so understanding protein function is essential for fully interpreting genomic data.
My experience with GPS strengthened my interest in connecting genomics with protein structure and function. Tools such as AlphaFold have revolutionised this field by allowing us to predict protein structures with high accuracy. This helps us understand how gene mutations affect protein function, including those that cause antimicrobial resistance and vaccine evasion.
For example, by studying the structure of resistance enzymes that render drugs ineffective, we can better understand how bacteria inactivate antibiotics and potentially predict which drugs will remain effective. Ultimately, understanding proteins helps us answer important questions about disease mechanisms and treatments.
What is unique about working on protein resources at EMBL-EBI?
EMBL-EBI provides a uniquely collaborative environment for developing and maintaining global biological data resources. Resources like UniProt are developed through international partnerships, including collaborations with the Swiss Institute of Bioinformatics (SIB) and the Protein Information Resource (PIR).
This global collaboration allows us to combine expertise from different scientific disciplines and ensure that protein knowledge is accurate, comprehensive, and accessible to researchers worldwide.
As a team leader, what are your priorities for the Protein Function Content Team?
One of my key priorities is to strengthen protein function annotation in bacteria, particularly those relevant to infectious diseases and antimicrobial resistance. Improving these resources will help researchers better understand bacterial biology and support the development of new treatments and vaccines.
I also want to ensure that our team continues to produce high-quality curated datasets while incorporating new technologies such as AI-assisted annotation. Maintaining accuracy and reliability while scaling up annotation will be essential for supporting the growing volume of biological data.
How is your team integrating AI?
We are working closely with the Protein Function Development team at EMBL-EBI and with international partners to develop AI-based approaches for protein function annotation.
For example, we are using a high-quality, manually curated protein database called UniProtKB as a training dataset for the Protein Natural Language Model (ProtNLM). ProtNLM was developed in partnership with Google DeepMind and uses deep learning to predict protein function.
These approaches allow us to combine expert knowledge with AI to improve the scale, speed, and consistency of protein annotation.
What do you enjoy doing outside of research and bioinformatics?
Outside of work, I enjoy sports, especially racket sports such as badminton, tennis, and padel. I also enjoy swimming. Sport helps me to unwind, stay active, and maintain a healthy balance between work and personal life.
I’m also lucky to share my life with my husband, Jimmy Lee, a senior data scientist at the Wellcome Sanger Institute, who is my source of energy and encouragement. Our daily scientific discussions inspire me and make science a part of both my work and life.
