This repository contains a collection of Jupyter Notebooks, which can be used to teach pharmaceutical and chemistry students the basics of Deep Learning. No prior coding knowledge is required. The article introducing this repository can be found here: https://doi.org/10.1002/ardp.202200628 and was written by Janosch Menke, Samuel Homberg and Oliver Koch.
This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.
This work was funded by the "Apotheker Stiftung Westfalen-Lippe"
-
Goolge Colab
The easiest way to use the Notebooks is to open them in Google Colab. The only thing needed is a Google Account. You can open a Juypter Notebook by simply clicking on a button in the table below. All notebooks will work out-of-the-box. -
Local Installation
If you do not want to run the notebooks through a Google service, you can also setup your own local Python environment. We provide an instruction on how to do this. Like with Colab all notebooks will work straight away, as soon as the local installation has been completed.
| Notebook | English | German |
|---|---|---|
| 01. Introduction to Jupyter |  |
|
| 02. Introduction to Python | |
|
| 03. Cheminformatics & RDKit | |
|
| 04. Linear Regression | |
|
| 05. Data Science | |
|
| 06. Linear Algebra | |
|
| 07. Your first Neural Network | |
|
| 08. PyTorch | |
|
| 09. Convolutional Neural Network | |
|
| 10. Transfer Learning | |
|
| 11. Recurrent Neural Networks | |
|
| 12. Autoencoders | |
|
| 13. Graph Neural Networks | |
|
| 14. Summary | |
|
We want to point out that these notebooks are, on their own, not sufficient to properly convey the knowledge and teach students about deep learning. Instructors need to prepare their own accompanying lectures. It is also important to mention that these notebooks are not designed to bring students to a level where they are able to train neural networks without any aid. Rather, the notebooks are designed to teach students the theoretical concepts to understand neural networks through code completion. We believe, as explained in more detail in the paper, that the theory bheind neural networks is easy to understand. But learning about them, is difficult as it requieres a solid understanding of a programming language. So students would get stuck on syntactical problems posed by the programming language rather than the theory behind neural networks.
We hope that these notebooks can be a starting point for others to expand on or contribute to. Everyone is free to adapt this repository (in accoradance with the above mentioned license).
| Name | Source |
|---|---|
| MNIST | LeCun, Y., Bottou, L., Bengio, Y., & Haffner, P. (1998). Gradient-based learning applied to document recognition. Proceedings of the IEEE, 86(11), 2278-2324. |
| BBBP | Martins, I. F., et al. (2012) A Bayesian approach to in silico blood-brain barrier penetration modeling. Journal of Chemical Information and Modeling, 52(6), 1686-1697. |
| Pneumonia | Kermany, D., Zhang, K., Goldbaum, M. (2018), Large Dataset of Labeled Optical Coherence Tomography (OCT) and Chest X-Ray Images, Mendeley Data, V3, doi: 10.17632/rscbjbr9sj.3 Kermany, D. S., Goldbaum, M., Cai, W., Valentim, C. C., Liang, H., Baxter, S. L., ... & Zhang, K. (2018). Identifying medical diagnoses and treatable diseases by image-based deep learning. Cell, 172(5), 1122-1131. |
| Cats & Dogs | Parkhi, O. M., Vedaldi, A., Zisserman, A., & Jawahar, C. V. (2012). Cats and dogs. In 2012 IEEE conference on computer vision and pattern recognition (pp. 3498-3505). IEEE. |
| GDB 11 | Fink, T., & Reymond, J. L. (2007). Virtual exploration of the chemical universe up to 11 atoms of C, N, O, F: assembly of 26.4 million structures (110.9 million stereoisomers) and analysis for new ring systems, stereochemistry, physicochemical properties, compound classes, and drug discovery. Journal of Chemical Information and Modeling, 47(2), 342-353. |
TeachOpenCADD A collection of notebooks covering a wide range of topics related to cheminformatics and data science, like collecting and cleaning molecular data in Python, but also more advanced topics like Docking.