Big Fantastic Virus Database (BFVD) - A Repository of Predicted Viral Protein Structures
Author Information
Author(s): Kim Rachel Seongeun, Levy Karin Eli, Mirdita Milot, Chikhi Rayan, Steinegger Martin
Primary Institution: Seoul National University
Hypothesis
The study aims to create a comprehensive database of predicted viral protein structures to enhance viral research.
Conclusion
BFVD provides a unique and extensive collection of viral protein structures, significantly improving the availability of resources for viral research.
Supporting Evidence
- BFVD contains 351,242 predicted viral protein structures.
- Over 62% of BFVD entries show low structural similarity to existing databases.
- BFVD can be queried using Foldseek and UniProt labels.
Takeaway
The BFVD is like a big library of virus protein shapes that helps scientists understand how viruses work better.
Methodology
The BFVD was constructed by predicting protein structures from viral sequences using ColabFold and improving predictions through homology searches.
Potential Biases
The reliance on computational predictions may introduce biases in the structural accuracy of some entries.
Limitations
BFVD may contain singletons that are less reliable due to shallow multiple sequence alignments.
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website