TeachOpenCADD

A teaching platform for computer-aided drug design (CADD) using open source packages and data.

If you use TeachOpenCADD in a publication, please cite us! If you use TeachOpenCADD in class, please include a link back to our repository.

In any case, please star (and tell your students to star) those repositories you consider useful for your learning/teaching activities.

Description

Figure adapted from Figure 1 in the TeachOpenCADD publication (D. Sydow et al., J. Cheminformatics, 2019).

Open source programming packages for cheminformatics and structural bioinformatics are powerful tools to build modular, reproducible, and reusable pipelines for computer-aided drug design (CADD). While documentation for such tools is available, only few freely accessible examples teach underlying concepts focused on CADD applications, addressing especially users new to the field.

TeachOpenCADD is a teaching platform developed by students for students, which provides teaching material for central CADD topics. Since we cover both the theoretical as well as practical aspect of these topics, the platform addresses students and researchers with a biological/chemical as well as a computational background.

Each topic is covered in an interactive Jupyter Notebook, using open source packages such as the Python packages rdkit, pypdb, biopandas, nglview, and mdanalysis (find the full list here). Topics are continuously expanded and open for contributions from the community. Beyond their teaching purpose, the TeachOpenCADD material can serve as starting point for users’ project-directed modifications and extensions.

New edition: we have extended the TeachOpenCADD platform with 6 notebooks introducing deep learning and its application to CADD related topics.

Get started

If you can't wait and just want to read through the materials, please go to the read-only version here.

You can run the TeachOpenCADD talktorials either in the cloud for an instant start or locally for a full development experience.

Run Online

The fastest way to explore is via Google Colab. No installation is required.

• Navigate to the Talktorials list below in Open in Google Colab section.
• Click the notebook URL on the title column to launch the tutorial directly in your browser.

---

Run Locally

To set up the project on your machine, follow these steps:

1. Clone the repository

Open your terminal and clone the repository:

git clone https://github.com/volkamerlab/teachopencadd.git
cd teachopencadd

2. Execute notebooks

The inside the main directory run the following command, please make sure you have conda installed.

python main.py -t T001

Here T001 is the ID referring to the talktorial to execute.

It will take couple of minutes if you're running a particular notebook for the very first time. First, internally it will create a conda virtual environment and install all the necessary python packages need, you can see the logs in the terminal. Then the notebook will launch in your browser.

Open in Google Colab

Talktorial	Title
T001	Compound data acquisition (ChEMBL)
T002	Molecular filtering: ADME and lead-likeness criteria
T003	Molecular filtering: unwanted substructures
T004	Ligand-based screening: compound similarity
T005	Compound clustering
T006	Maximum common substructure
T007	Ligand-based screening: machine learning
T008	Protein data acquisition: Protein Data Bank (PDB)
T009	Ligand-based pharmacophores
T010	Binding site similarity and off-target prediction
T011	Querying online API webservices
T012	Data acquisition from KLIFS
T013	Data acquisition from PubChem
T014	Binding site detection
T015	Protein ligand docking
T016	Protein-ligand interactions
T017	Advanced NGLview usage
T018	Automated pipeline for lead optimization
T019	Molecular dynamics simulation
T020	Analyzing molecular dynamics simulations
T021	One-Hot Encoding
T022	Ligand-based screening: neural networks
T023	What is a kinase?
T024	Kinase similarity: Sequence
T025	Kinase similarity: Kinase pocket (KiSSim fingerprint)
T026	Kinase similarity: Interaction fingerprints
T027	Kinase similarity: Ligand profile
T028	Kinase similarity: Compare different perspectives
T033	Molecular representations
T034	RNN-based molecular property prediction
T035	GNN-based molecular property prediction
T036	An introduction to E(3)-invariant graph neural networks
T037	Uncertainty estimation
T038	Protein Ligand Interaction Prediction

TeachOpenCADD KNIME workflows

If you prefer to work in the context of a graphical interface, talktorials T001-T008 are also available as KNIME workflows. Questions regarding this version should be addressed using the "Discussion section" available at this post. You need to create a KNIME account to use the forum.

About TeachOpenCADD

• Contact
• Acknowledgments
• Citation
• License
• Funding

External resources

Please refer to our TeachOpenCADD website to find a list of external resources:

• External packages and webservices that are used in the TeachOpenCADD material
• Further reading material on Python programming, cheminformatics, structural bioinformatics, and more.