SCMMTP

SCMMTP predicts and characterizes membrane transport proteins using a scoring card method based on dipeptide composition to support structural and functional genomics.

Key Features:

• Scoring Card Method (SCM): SCMMTP employs the scoring card method (SCM) using dipeptide composition for identification and characterization of membrane transport proteins (MTPs).
• Dataset Utilization: Developed using a dataset of 900 MTPs and 660 non-MTPs split into a training set of 1,380 proteins and an independent test set of 180 proteins.
• Propensity Score Estimation: Generates propensity scores for amino acids and dipeptides that indicate likelihood of inclusion in MTPs.
• Accuracy Levels: Reports a training accuracy of 83.81% and a test accuracy of 76.11%, compared with an SVM using PSSM test accuracy of 80.56%.
• Application Across Datasets: Applied to human transmembrane proteins, photosynthetic protein datasets, and a comprehensive human protein database.
• Structural Insights: Identifies MTPs as α-helix rich and highlights the involvement of residues with low hydration energy, suggesting release of hydrated water molecules from pore regions after substrate filtering.

Scientific Applications:

• Membrane transport mechanism research: Enables analysis of membrane transport mechanisms by identifying and characterizing MTPs.
• Novel MTP prediction and experimental prioritization: Predicts novel MTPs based on dipeptide composition for experimental validation.
• Structural and functional annotation: Uses propensity scores to infer structural features and functional roles of MTPs in genomics and proteomics studies.

Methodology:

Applies the scoring card method (SCM) on dipeptide composition, computes propensity scores for amino acids and dipeptides, and was trained/tested on a dataset of 900 MTPs and 660 non-MTPs (training 1,380; independent test 180), with performance compared to SVM using PSSM.