iSNO-PseAAC

iSNO-PseAAC predicts S-nitrosylation (SNO) sites in protein sequences, identifying cysteine residues modified by nitric oxide to support study of this posttranslational modification.

Key Features:

• Integration of PSAAP into PseAAC: Combines position-specific amino acid propensity (PSAAP) with the pseudo amino acid composition (PseAAC) framework for feature representation of sequence context.
• Conditional Random Field (CRF) model: Uses a conditional random field algorithm for structured prediction of nitrosylated sites within protein sequences.
• Benchmark dataset: Employs a curated dataset comprising 731 SNO sites and 810 non-SNO sites with proteins filtered to reduce homology by excluding high pairwise sequence identity.
• Cross-validation performance: Demonstrates an overall cross-validation success rate exceeding 90% on independent datasets.

Scientific Applications:

• S-nitrosylation site identification: Predicts potential S-nitrosylation sites on cysteine residues in proteins for experimental prioritization.
• Protein function and regulation studies: Supports investigations of how S-nitrosylation affects protein function and regulatory mechanisms.
• Cellular signaling analysis: Aids in elucidating roles of S-nitrosylation in cellular signaling pathways.
• Drug discovery and therapeutic targeting: Assists identification of SNO sites that may inform development of interventions targeting posttranslational modifications.

Methodology:

Feature extraction via PseAAC incorporating PSAAP, model training and prediction using a conditional random field (CRF), and evaluation on a curated benchmark of 731 SNO and 810 non-SNO sites filtered for low pairwise sequence identity with cross-validation assessment (>90% on independent datasets).