ResSUMO

ResSUMO predicts lysine sumoylation sites from protein sequences to identify post-translational modification loci relevant to proteomics and disease studies.

Key Features:

• Protein sequence input: Uses protein sequence data as the input for site prediction.
• ZScale encoding: Applies the ZScale encoding scheme to convert amino acids into sequence-based and physicochemical feature vectors.
• One-dimensional convolutional neural network (1-D CNN): Employs a 1-D CNN architecture tailored for sequence data to detect patterns associated with SUMOylation.
• Residual structure integration: Incorporates residual connections within the neural network to mitigate degradation in deeper architectures.
• Performance metrics: Reported an area under the ROC curve of approximately 0.80 in comparative evaluations against traditional machine learning models and other CNN approaches.

Scientific Applications:

• Proteomics research: Predicts lysine SUMOylation sites to inform functional and large-scale proteomic analyses.
• Post-translational modification studies: Supports investigation of the roles of SUMOylation in cellular processes and disease mechanisms.

Methodology:

Constructed and compared 16 classifiers by integrating four different algorithms with four encoding features selected from a pool of 11 sequence-based or physicochemical features, and employed ZScale encoding, a 1-D CNN architecture, and residual connections.