iACetyP

iACetyP predicts acetylated proteins from sequence to identify protein acetylation, a post-translational modification involving the addition of an acetyl group from acetic acid, using conservation-derived features (gray system model and K-nearest neighbors scores), functional domain and subcellular localization annotations, and Relief-based feature selection, and was validated by 5-fold cross-validation on three datasets achieving mean accuracy 77.10%, Matthew's correlation coefficient 0.5457, and AUC 0.8389.

Key Features:

• Feature Extraction: Employs conservation information via the gray system model and K-nearest neighbors (KNN) scores and incorporates functional domain annotation and subcellular localization.
• Feature Selection: Uses the Relief algorithm to refine and select the most relevant predictive features.
• Validation and Performance: Evaluated by 5-fold cross-validation on three datasets with mean accuracy 77.10%, Matthew's correlation coefficient 0.5457, and AUC 0.8389.
• Input Format: Processes protein sequences provided in FASTA format.

Scientific Applications:

• Identification of acetylated proteins: Distinguishes acetylated from non-acetylated proteins to support studies of protein acetylation.
• Complement to experimental techniques: Complements high-throughput mass spectrometry and can guide experimental validation of acetylation and other PTMs.
• Elucidation of PTM mechanisms: Aids investigation into the mechanisms underlying protein modifications and acetylation-related regulation.

Methodology:

Feature extraction using sequence conservation via the gray system model and K-nearest neighbors (KNN) scores, integration of functional domain annotation and subcellular localization, feature selection with Relief, and validation by 5-fold cross-validation on three datasets.