Related Experiment Video
Updated: Jan 8, 2026

Author Spotlight: A Computational Approach to Decipher Amino Acid Preferences in Multispecific Protein-Protein Interactions
Published on: January 26, 2024
Assessing the validity of leucine zipper constructs predicted by AlphaFold
Isobel Mitic1,2, Keiran Rowell3, Thomas Litfin3
1Department of Molecular Medicine, School of Biomedical Sciences, University of New South Wales, Sydney, New South Wales, Australia.
Artificial intelligence (AI) accurately predicts protein structures, including leucine zipper dimers, but may overestimate confidence for improbable interactions. This highlights AI
Area of Science:
- Structural biology
- Computational biology
- Genetics
Background:
- AP-1 transcription factors regulate cellular pathways through dimerization via leucine zipper domains.
- Dimerization specificity is determined by transcription factor affinity and concentration, controlling gene expression.
- AI-driven protein structure prediction offers new avenues for studying protein interactions.
Purpose of the Study:
- To investigate AlphaFold's ability to model leucine zipper domains and predict dimer interfaces.
- To assess AlphaFold's accuracy in differentiating between probable and improbable protein-protein interactions.
- To evaluate the capabilities and limitations of AI in structural biology.
Main Methods:
- Utilized AlphaFold2 and AlphaFold3 for AI-based protein structure prediction.
- Tested AlphaFold's performance on the Fos-Jun AP-1 dimer model.
- Analyzed over 2000 experimentally validated human leucine zipper sequences.
Main Results:
- AlphaFold successfully predicted highly confident leucine zipper dimers.
- The AI model predicted the formation of the FosB homodimer, despite known electrostatic repulsion in vivo.
- Demonstrated AlphaFold's potential for high-confidence but potentially inaccurate structure predictions.
Conclusions:
- AlphaFold can model leucine zipper domains with high confidence.
- AI predictions may not always align with experimental biological constraints, such as electrostatics.
- Understanding AI limitations is crucial for accurate structural biology applications.
Related Concept Videos
Conserved Binding Sites
Binding sites are often located in large pockets, and if their location on a protein’s surface is unknown, it can be predicted using various approaches. The energetic method computationally...
Protein Folding Quality Check in the RER
Protein Folding
Protein Folding
Protein Structure Is Critical to Its Biological Function
Proteins perform a wide range of biological functions such as catalyzing chemical reactions, providing...
Conservation of Protein Domains Over Different Proteins
A limited set of protein domains often duplicate and recombine during evolution. These domains can be organized in different combinations to...
Protein Complexes with Interchangeable Parts
The SCF ubiquitin ligase is a protein complex of five individual proteins. This complex attaches ubiquitin to other target proteins to mark them for degradation. In order...

