Introduction

This repository contains a set of notebooks (organized into seperate topics by lab) for an introduction to cheminformatics in the Python programming language. Introductions to programming using the Python programming language are included. The notebooks rely on a set of both dependecies Python and other packages listed in the environmental.yml. Thus, installing a Python environment using conda is required. Conda installation can be done as described on the conda website. Some notebooks are taken from the incredible Teach Open Computer-aided Drug Design course. However, notebooks in this repository are geared to noive programmers.

Running the notebooks on a sever

These noteboks can be run a server by clicking the Binder tag above. However, all changes to the notebooks (e.g., your work) and therefore is only recommended as a temporary solution.

Running the notebooks locally

After conda is properly installed, a new conda environment for these notebooks can be created by the command:

conda env create -f environment.yml

The environment can then be activated by:

conda activate intro-chem

And the notebooks can be opened via:

jupyter lab

or

jupyter notebook

Overview

1. Course Introduction
   #TODO: This is empty

2. Introduction to chemical structure annotations– 2D &3D
   Teach structure annotations: ChemBioOffice, SMILE, SDF and etc. First homework assignment #TODO: This is empty

3. Basic Python, Pandas and matplotlib
   Introduce the basics of programming using the Python language. Introduce popular data science and visualization packages, Pandas and matplotlib.

4. Advanced python and RDKit
   A brief introduction to Objects in Python. Introduction to the cheminformatics Python library RDKit.

5. Molecular Descriptors
   Generating chemical descriptors - explores the basics of generting molecular descriptors and chemical fingerprints.

6. Chemical Similarity
   The basics of chemical similarities using different metrics and descriptor spaces.

7. Machine Learning Part I - Unsupervised ML
   The basics of unsupervised machine learning, including principal component analysis and chemical clustering.

8. Machine Learning Part I - QSAR
   The basics of supervised machine learning and developing quantitative structure activity relationship (QSAR) models. Develop classification and regression models on a set of benzodiazipines.

9. Data science in chemistry
   Learning about web retrieval services and how to access chemistry databases via programmatic interfaces.

10. Predictive modeling
    Develop your own QSAR models and make predictions.

11. Pharmacophores
    Generating pharmacophores using a ligand based approach for EGFR receptor

12. Deep Learning
    Develop a deep learning model using data provided from the NICEATM acute oral toxicity challenge.