The multilayer Perceptron algorithm trained on Proteomics dataset of Parkinson disease for screening the new potential proteome.
Genome-wide association studies (GWAS) have provided a deeper understanding of the genetic basis of complex diseases, including Parkinson's disease (PD). However, there is still a need to identify new disease gene associations for PD. This research has the potential to generate new and improved hypotheses about the pathogenesis of PD.
This repository contains a parkinson-explorer that was trained on a dataset of 235 putative PD-associated proteins and 250 random proteins. The classifier achieved the following evaluation metrics:
- Precision: 95%
- Recall: 93%
- F1: 94%
- ROC: 98%
The classifier was also used to screen a neurodegenerative gene set consisting of 400 proteins to predict a new potential proteome consisting of 38 proteins. The new proteome plays a pivotal role in and around the synapse, performing molecular functions like RNA-binding, growth factors, and neuropeptides; playing or inserting an effect over major processes in and around mRNA processing, mRNA splicing, and neurogenesis.
To install the parkinson-explorer, you will need to have Python 3 and the following Python packages:
- numpy
- pandas
- scikit-learn
- jupyter
You can install the dependencies using the following command:
pip install numpy pandas scikit-learn jupyter
Once the dependencies are installed, you can clone the repository using the following command:
git clone https://github.com/aysanraza/parkinson-explorer.git
To use the carcinoma classifier, you will need to open the parkinson-explorer.ipynb file in Jupyter Notebook.
- 0.1
- Initial Release
This project is licensed under the MIT license - see the LICENSE.md file for details
- Ahsan Raza