iLoc-Cell

iLoc-Cell predicts the subcellular localization of human proteins, including proteins that occupy multiple distinct subcellular sites.

Key Features:

• Multiplex Localization Prediction: Identifies both single-location and multiple-location assignments for human proteins, addressing proteins that occupy two or more subcellular sites.
• Comprehensive Coverage: Covers 14 subcellular sites: centrosome, cytoplasm, cytoskeleton, endoplasmic reticulum, endosome, extracellular space, Golgi apparatus, lysosome, microsome, mitochondrion, nucleus, peroxisome, plasma membrane, and synapse.
• High Accuracy: Achieved an overall success rate of 76% in jackknife cross-validation on datasets containing proteins assigned to two, three, or four locations, with subsets constrained to no more than 25% pairwise sequence identity.
• Independent Validation: Performance was confirmed using two independent datasets, with success rates reported as superior to existing predictors capable of handling multi-location proteins.

Scientific Applications:

• Functional Genomics: Facilitates assignment of proteins to cellular compartments to infer functional roles within cellular contexts.
• Protein Engineering and Drug Design: Provides subcellular localization information useful for targeting or modifying proteins for therapeutic or engineering purposes.
• Cell Biology Research: Supports studies of cellular dynamics and interactions by mapping proteins to their respective cellular sites.

Methodology:

Benchmark tests employed jackknife cross-validation on datasets of proteins annotated with two, three, or four locations (with ≤25% pairwise sequence identity within subsets) and validation was performed on two independent datasets.