AgMata

AgMata predicts protein beta-aggregation regions from amino acid sequences to identify aggregation-prone residues and evaluate the effects of point mutations.

Key Features:

• Unsupervised approach: Employs an unsupervised methodology to identify aggregation-prone regions within protein sequences without prior training on labeled datasets.
• Statistical potentials: Uses a generalized definition of statistical potentials that incorporate biophysical information to score aggregation propensity.
• Benchmark performance: Demonstrated superior performance compared to state-of-the-art methods across two distinct benchmarks.
• Ataxin-3 case study: Identified aggregation-prone residues in human ataxin-3 (implicated in Machado-Joseph disease) that share similar structural environments.
• Mutational prediction and validation: Predicts effects of point mutations on aggregation propensity, validated by in vitro mutagenesis experiments on wild-type ataxin-3.

Scientific Applications:

• Protein misfolding disease research: Analyzing structural determinants of beta-aggregation in disorders such as Machado-Joseph disease.
• Therapeutic development: Informing strategies to target or modulate aggregation in disease contexts.
• Protein engineering and experimental design: Guiding mutagenesis experiments by predicting mutational impacts on aggregation propensity.
• Drug discovery: Supporting target selection and lead optimization by identifying aggregation-prone regions.

Methodology:

Applies an unsupervised computational approach based on a generalized definition of statistical potentials that integrate biophysical information to identify aggregation-prone regions.