https://github.com/frknbm/protein_secondary_structure_prediction_and_classification

Protein Secondary Structure Prediction and Classification
https://github.com/frknbm/protein_secondary_structure_prediction_and_classification

Last synced: about 2 months ago
JSON representation

Protein Secondary Structure Prediction and Classification

• Host: GitHub
• URL: https://github.com/frknbm/protein_secondary_structure_prediction_and_classification
• Owner: frknbm
• Created: 2024-01-17T14:13:49.000Z (12 months ago)
• Default Branch: main
• Last Pushed: 2024-01-17T14:24:45.000Z (12 months ago)
• Last Synced: 2024-01-17T22:44:28.838Z (12 months ago)
• Language: Jupyter Notebook
• Size: 1.89 MB
• Stars: 0
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
# Project Title: Protein Secondary Structure Prediction and Classification

## Description

This project involves predicting the secondary structures of proteins using different digitization methods and classification algorithms. The dataset contains protein IDs, sequences, and their corresponding secondary structures. Two distinct digitization methods, EIIP and CPNR, were applied, and classification processes were executed.

### EIIP Method Classification

#### Naive Bayes

- Protein secondary structures were classified using the Naive Bayes classifier.

- Performance metrics (accuracy, F1-score, precision, recall) were calculated.

- Classification results were visualized, and a confusion matrix was generated.

#### DVM

- Protein secondary structures were classified using the DVM classifier.

- Performance metrics (accuracy, F1-score, precision, recall) were calculated.

- Classification results were visualized, and a confusion matrix was generated.

### CPNR Method Classification

#### LSTM

- Protein secondary structures were classified using the Long Short-Term Memory (LSTM) model.

- Performance metrics (accuracy, F1-score, precision, recall) were calculated.

- Classification results were visualized, a confusion matrix was generated, and an ROC curve was plotted.

- ROC curves for each class were provided.

#### RNN

- Protein secondary structures were classified using the Recurrent Neural Network (RNN) model.

- Performance metrics (accuracy, F1-score, precision, recall) were calculated.

- Classification results were visualized, a confusion matrix was generated, and an ROC curve was plotted.

- ROC curves for each class were provided.

## Running the Project

Download the project files to your computer. Install the required libraries, and run the project files to observe the results.

## Used Libraries

- NumPy

- Pandas

- Scikit-learn

- Matplotlib

- TensorFlow (for LSTM and RNN)

## Contact

If you have any questions or feedback regarding the project, please contact me at [[email protected]],[https://www.linkedin.com/in/furkan-bayram-7b3499220].