iSMP-Grey

iSMP-Grey predicts secretory versus non-secretory proteins in malaria parasites by leveraging evolutionary information encoded in position-specific scoring matrices (PSSMs) to aid identification of potential drug targets.

Key Features:

• Classification task: Predicts secretory and non-secretory proteins in malaria parasites.
• Evolutionary information (PSSMs): Incorporates sequence evolution information via position-specific scoring matrices to capture evolutionary constraints on protein sequences.
• Pseudo Amino Acid Composition and grey system model: Uses a 60-dimensional feature vector that integrates the general form of PseAAC with sequence-evolution information through a grey system model, effective for handling incomplete or uncertain data.
• Validation and performance: Achieved a 94.8% overall success rate in identifying secretory proteins as measured by the jackknife test.

Scientific Applications:

• Drug target identification: Prioritizes secreted proteins of malaria parasites as potential antimalarial drug targets.
• Investigation of parasite survival mechanisms: Facilitates identification of proteins that mediate parasite survival within host erythrocytes.
• Candidate prioritization for experimental validation: Provides computational predictions to accelerate identification of secretory proteins and reduce reliance on labor-intensive experiments.

Methodology:

Uses position-specific scoring matrices (PSSMs) for evolutionary information; constructs a 60-dimensional feature vector combining the general form of pseudo amino acid composition (PseAAC) with sequence-evolution information via a grey system model; and evaluates performance using the jackknife test (reported 94.8% success rate).