Related Experiment Video
Updated: Aug 3, 2026
11:39
Assessing Autophagic Flux by Measuring LC3, p62, and LAMP1 Co-localization Using Multispectral Imaging Flow Cytometry
Published on: July 21, 2017
Predicting apoptosis protein subcellular localization by integrating auto-cross correlation and PSSM into Chou's
1School of Mathematics and Statistics, Xidian University, Xi'an 710071, PR China.
Journal of Theoretical Biology
|September 5, 2018
Summary
Predicting apoptosis protein subcellular localization is challenging. The novel MACC-PSSM model integrates Moran autocorrelation and cross-correlation with PSSM, achieving high prediction accuracies of 84.9% and 90.5% on benchmark datasets.
Area of Science:
- Computational Biology
- Bioinformatics
- Proteomics
Background:
- Predicting subcellular localization of apoptosis proteins is crucial for understanding cell death pathways.
- Current methods primarily rely on protein primary sequences, facing limitations in accuracy.
- Apoptosis protein localization impacts cellular function and disease mechanisms.
Purpose of the Study:
- To develop a novel computational model for accurately predicting apoptosis protein subcellular localization.
- To integrate advanced feature extraction techniques for improved prediction performance.
- To evaluate the proposed model against established benchmark datasets.
Main Methods:
- Developed the MACC-PSSM model by combining Moran autocorrelation and cross-correlation with Position-Specific Scoring Matrices (PSSM).
- Constructed a high-dimensional feature vector (3600-dimensional) for apoptosis protein representation.
- Applied Principal Component Analysis (PCA) for feature selection, reducing dimensionality to 210 features.
- Utilized Support Vector Machine (SVM) as the classification algorithm.
- Validated the model using jackknife cross-validation on ZW225 and CL317 datasets.
Main Results:
- The MACC-PSSM model achieved high prediction accuracies.
- Overall prediction accuracy reached 84.9% for the ZW225 dataset.
- Overall prediction accuracy reached 90.5% for the CL317 dataset.
- Demonstrated competitive performance compared to existing methods.
Conclusions:
- The MACC-PSSM model offers a robust and accurate approach for predicting apoptosis protein subcellular localization.
- The integration of autocorrelation features and PSSM enhances prediction capabilities.
- MACC-PSSM serves as a valuable tool for bioinformatics research and drug discovery.
