Protein aggregation is the phenomenon which occurs when misfolded or unfolded protein physically binds together and can cause the development of various amyloidosis diseases. The goal of this study was to construct surrogate models for predicting protein aggregation using data-driven methods with two types of databases. This study suggests which approaches is more effective to predict protein aggregation depending on types of descriptors and database.
-
Computational approach
- Feature-based Model
- Graph-based Model
-
Experimental approach
- Feature-based Model