AmyloGram reveals amyloidogenic potential in stroke thrombus proteomes

Area of Science:

• Biochemistry
• Proteomics
• Computational Biology

Background:

• Amyloidogenic proteins are implicated in various diseases.
• Their role in thrombi is a recent discovery.
• Computational tools for predicting amyloid formation in thrombus proteomes are needed.

Purpose of the Study:

• To evaluate the computational tool AmyloGram for predicting amyloid-forming potential.
• To analyze the amyloidogenicity of proteins in human stroke thrombi.
• To investigate the potential involvement of amyloid mechanisms in stroke pathophysiology.

Main Methods:

• AmyloGram, a tool using n-gram sequence encoding and random forest classification, was employed.
• Analysis included 204 known amyloidogenic proteins from UniProt.
• Proteins from stroke thrombi (cardioembolic and atherothrombotic) and 83,567 human protein sequences were analyzed.

Main Results:

• AmyloGram accurately identified known amyloidogenic proteins, with most scoring above 0.7.
• Proteins within stroke thrombi consistently showed high AmyloGram scores (>0.7), indicating significant amyloidogenic potential.
• A majority of unannotated proteins also exhibited high amyloidogenicity scores.

Conclusions:

• AmyloGram is a reliable tool for identifying amyloid-forming proteins.
• Stroke thrombi are significantly enriched with proteins possessing high amyloidogenic potential.
• Findings support the hypothesis that amyloid-related mechanisms are involved in stroke thrombus formation, meriting further study.