AlphaFold Prediction of Protein–Protein Interactions in the Flaviviridae Proteomes
The family <i>Flaviviridae</i> is divided into flaviviruses, hepaciviruses and pestiviruses. Its members infect a wide range of organisms, from insects to humans, and share a similar genome organization where proteins require sequential cleavage from a single translated polyprotein. Despite decades of study, the structures of some non-structural (NS) membrane proteins, or details of their protein-protein interactions (PPIs), are still unclear. Since AlphaFold (AF) can be used to predict interactions between protein domains using Predicted Alignment Error (PAE) score plots, we hypothesized that AF-predicted interactions between domains of complete viral polyproteins can represent PPIs if these interactions are retained once the different proteins are sequentially cleaved. We complemented this approach using AF predictions involving all independent separate protein sequences, instead of using a single polyprotein. We found that most of these PPIs have already been reported experimentally, which validates the use of AF in this context, but not all of these PPIs have been characterized from a structural perspective. Thus, we propose that AF provides testable hypotheses regarding residues involved in these PPIs, and that comparison of the three genera in this family may provide much needed clues to the function of these proteins during the viral life cycle.
Authors
- Jaume Torres (ORCID: https://orcid.org/0000-0002-1282-9727)
- Wahyu Surya (ORCID: https://orcid.org/0000-0001-9240-371X)
- Joshua‐Daniel Goh
- Caleb Ponniah
Institutions
- Nanyang Technological University (SG)
Publication Details
- Journal
- International Journal of Molecular Sciences
- Published
- 2025-10-19
- DOI
- https://doi.org/10.3390/ijms262010159
- Citations
- 1
- Primary Topic
- Machine Learning in Bioinformatics
- Type
- article
- Field-Weighted Citation Impact
- 0.81
Funders
- Ministry of Education, India