Skip to content
Stockholm University · SciLifeLab

AI and structural bioinformatics for proteins, RNA, and membranes

The Elofsson Lab develops computational methods to understand how proteins and RNA fold, move, interact, evolve, and function, combining machine learning, structural bioinformatics, and large-scale biological data.

Explore research Join the lab Publications
Protein interactions Protein evolution Membrane proteins Machine learning
Research snapshot
Structure-aware analysis of proteins, domains, and interactions
Structure-aware maps of domains and interactions across molecular systems.
Recruitment
Open to strong PhD and postdoc enquiries
202+
Publications listed in the current CV
162.9 MSEK
Total current grants
4.2 MSEK
VR 2025 project on protein domain architecture evolution
3
Postdocs, two in the lab and one shared with EMBL-EBI

Founding and strategic funders

Major support from KAW, DDLS, EU and Vetenskapsradet (VR).

Knut and Alice Wallenberg Foundation Vetenskapsradet European Commision DDLS

News

Latest social updates from Bluesky.

Research areas

The group combines structural bioinformatics, machine learning, and evolutionary analysis to understand proteins, complexes, membranes, and molecular systems.

Protein dynamics and conformational change

Equivariant models that predict how a protein moves directly from its structure, and generative methods that design proteins with prescribed flexibility.

RNA structure and ion-dependent folding

Why deep learning still struggles with RNA: shallow homology and neglected ions, and what it takes to fix both.

Protein structure and complex prediction

From consensus and quality assessment methods to AlphaFold-era modelling of protein interactions, assemblies, and antibody-antigen recognition.

Protein evolution and domain architecture

How domains change, duplicate, and rearrange across the tree of life, now with structure-aware analysis using AlphaFoldDB, TED, and ECOD.

Membrane proteins, from topology to whole membranes

Topology and structure of membrane proteins, extended to multiscale models of whole membranes within CEMERAI.

AI for proteoforms and molecular interactions

Representation learning, graph methods, and protein language models for proteoforms, interactions, and design.

Visual snapshots from current work

Recent projects combine large-scale structural data, evolutionary analysis, and deep learning to address questions that range from molecular systems in the cell to the emergence of protein architectures across the tree of life.

Overview of domain analysis with AlphaFoldDB and TED
Structure-aware analysis of protein domains and their evolution.
Examples of domain architecture evolution
Domain architecture changes and structural variation across protein families.

About the lab

The lab led by Arne Elofsson is known for longstanding contributions in protein structure prediction, membrane protein topology, protein model quality assessment, and protein evolution. Today, this foundation is being extended into AlphaFold-era modelling, proteome-scale interaction and complex prediction, and AI-driven analysis of proteoforms and molecular systems.

8 PhD students
3 MSc students
3 Postdocs
2 Senior scientists
9 Alumni in faculty
29+ PhD graduates

Why join?

We offer projects at the interface of AI, protein science, and structural biology, with strong links to Stockholm University, SciLifeLab, and national computational infrastructure.