ngLOC

ngLOC predicts protein subcellular localization, assigning single or multiple subcellular locations for proteins across eukaryotic and prokaryotic species.

Key Features:

• Comprehensive Prediction Capabilities: ngLOC can identify multiple subcellular locations for proteins, predicting 11 distinct locations in plant and animal species, 4 locations in gram-positive bacterial datasets, and 5 locations in gram-negative bacterial datasets.
• High Prediction Accuracy: Overall prediction accuracy ranges from 89.8% to 91.4% across various species.
• Versatility Across Species: ngLOC is a generic method that can be trained using data from a wide range of species or classes, enabling applicability to diverse organisms.

Scientific Applications:

• Protein localization and functional inference: ngLOC supports studies in cell biology, molecular biology, and bioinformatics by predicting localization patterns including proteins that shuttle between cellular compartments, aiding functional and interaction analyses.

Methodology:

ngLOC employs an n-gram-based Bayesian classifier to analyze protein sequences, leveraging the statistical properties of sequence motifs (n-grams) to infer subcellular localizations and can be trained with diverse datasets from various species.