ngLOC: an n-gram-based Bayesian method for estimating the subcellular proteomes of eukaryotes

Area of Science:

• Bioinformatics
• Computational Biology
• Molecular Biology

Background:

• Accurate prediction of protein subcellular localization is crucial for understanding cellular functions.
• Existing methods may have limitations in handling proteins with multiple localization sites.

Purpose of the Study:

• To introduce ngLOC, a novel n-gram-based Bayesian classifier for predicting protein subcellular localization.
• To evaluate the performance of ngLOC for proteins localized to single and multiple organelles.
• To develop an enhanced ngLOC for estimating subcellular proteomes across diverse eukaryotic organisms.

Main Methods:

• Utilized an n-gram-based approach combined with Bayesian classification.
• Employed tenfold cross-validation for performance assessment.
• Extended the ngLOC method for large-scale proteome analysis.

Main Results:

• Achieved 89% accuracy for predicting single-organelle protein localization.
• Attained 82% accuracy for predicting multiple-organelle protein localization.
• Successfully applied an enhanced ngLOC to estimate subcellular proteomes in eight eukaryotic species.

Conclusions:

• ngLOC provides a robust and accurate method for predicting protein subcellular localization.
• The enhanced ngLOC is a valuable tool for comparative proteomic studies across eukaryotes.
• This work facilitates a deeper understanding of cellular organization and protein function.