Applied Machine Learning Class S23

Welcome to the repository for the Applied Machine Learning class taught by Tim Conrad. Throughout this course, we've tackled various bioinformatics challenges using both machine learning and deep learning techniques. Each project is accompanied by a scientific report, providing in-depth insights into our methodologies and findings.

General Project Structure

• Project_Name/
  • data/: Contains the dataset for the project
  • source/: Holds the source code
  • report/: Includes the scientific report for the project

Project Summaries

1. Project 1-2: Applied different machine learning methods to classify two classes of medications using metabolites dataset and visualized the results.
2. Project 3: Developed a machine learning model to automate somatic variant refinement.
3. Project 4: Utilized Random Forest classifier to distinguish between healthy and cancer samples.
4. Project 5: Built an eXtreme Gradient Boosting classifier for predicting colorectal cancer stages (adenoma to carcinoma) using gene expression data.
5. Project 6: Predicted cancer based on DNA methylation data.
6. Project 7: Explored proteomics datasets, converting LC-MS/MS information into machine-learning-friendly form using PyOpenMS library and binning feature engineering method.
7. Project 8: Applied three different embedding models from the pyKEEN package to perform multi-class link prediction for two different use-cases using a subset of the Hetionet dataset.

Feel free to explore the projects and reports for more details on our methodologies and findings.