The Fusarium Protein Toolkit (FPT) is a set of tools and datasets to explore the functions and structures of the Fusarium proteome. The resources are developed using MaizeGDB pipelines, enabling direct comparisons with maize datasets and tools. This approach allows for an optimized method of comparing proteins within a host/pathogen relationship.
Foldseek: Protein Structure Search
The AlphaFold predicted protein structure from Fusarium graminearum (15,911 proteins) and Fusarium verticillioides (17,356 proteins) were aligned against nine proteomes using the software Foldseek. The nine proteomes include six strains of Fusarium: F. graminearum, F. fujikuroi, F. oxysporum, F. proliferatum, F. solani, F. verticillioides. Four well-annotated proteomes were chosen as outgroups: Arabidopsis thaliana (Arabidopsis), Homo sapiens (Human), Saccharomyces cerevisiae (Budding yeast), Schizosaccharomyces pombe (Fission yeast). Try searching your gene or protein of interest in the box below!
PanEffect: Variant Effect Viewer
PanEffect is a JavaScript framework to explore variant effects across a pan-genome. The tool has two views that allows a user to (1) explore all possible amino acid substitutions and their variant effects for a reference genome (F. verticillioides or F. graminearum.), and (2) view the natural variation and their effects across the Fusarium pan-genome (created from 22 Fusarium species).
Predicted Effectors
This Fusarium Effector webpage features a curated collection of 290 effector proteins predicted for Fusarium graminearum. It includes a detailed table presenting this protein set and metadata explaining the annotation process. Additionally, the page offers direct links to the FoldSeek and PanEffect tools for further exploration and analysis.
AlphaFold is an artificial intelligence system by DeepMind (reference), and was used to create predictions based on the Fusarium graminearum (15,911 proteins) and Fusarium verticillioides (17,356 proteins) sequences. Enter a gene or protein below to view the predicted AlphaFold structure.
The protein structure is color-coded based on per-residue confidence scores (pLDDT):
Very high (avg pLDDT > 90) Confident (70 < avg pLDDT < 90) Low (50 < avg pLDDT < 70) Very low (avg pLDDT < 50)
ESMFold is an artificial intelligence system by MetaAI (reference), and was used to create predictions of Fusarium graminearum (15,911 proteins) and Fusarium verticillioides (17,356 proteins) sequences. Enter a gene or protein below to view the predicted ESMFold structure.
The protein structure is color-coded based on per-residue confidence scores (pLDDT):
Very high (avg pLDDT > 90) Confident (70 < avg pLDDT < 90) Low (50 < avg pLDDT < 70)
Very low (avg pLDDT < 50)
