Discriminating Membrane Proteins Using Machine Learning
Author Information
Author(s): Gromiha M Michael, Yabuki Yukimitsu
Primary Institution: Computational Biology Research Center (CBRC), National Institute of Advanced Industrial Science and Technology (AIST)
Hypothesis
Can machine learning techniques effectively discriminate between different classes of membrane proteins based on their amino acid compositions?
Conclusion
The study found that using amino acid occurrence improves the accuracy of discriminating membrane proteins compared to using amino acid composition.
Supporting Evidence
- The neural network method achieved a 5-fold cross-validation accuracy of 64%.
- Using amino acid occurrence improved the overall accuracy to 68%.
- The k-nearest neighbor method discriminated transporters with an accuracy of 85%.
Takeaway
This study shows how scientists can use computers to tell different types of proteins apart by looking at the building blocks they are made of.
Methodology
Different machine learning algorithms were applied to analyze amino acid compositions and occurrences in membrane proteins.
Limitations
The dataset may not represent all membrane proteins due to the limited number of protein structures analyzed.
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website