During this workshop I will show you how to build machine learning models for Computational Drug Discovery. More background on the subject may be found here
The first thing you will need to do is clone this repository, i.e. git clone [email protected]:jp-um/ARI5902_Research_Topics_In_Artificial_Intelligence.git
Note that this session assumes you have access to a Linux installation (ideally Ubuntu 24.04.1 LTS). You can use Oracle VirtualBox to create an Ubuntu Virtual Machine on your Windows/macOS (or you can dual boot, or even better, remove your OS and install Ubuntu directly on the bare metal).
For my session you will need docker installed (on Ubuntu). Docker is a clever bit of software which allows me to package all the libraries (and setup) you need all in one place. Docker can easily be installed (on Ubuntu, but also on other platforms) by following the official Docker guides.
You will need a ton of software (configuration, etc.), which I have conveniently installed and set-up for you and placed in a docker container (you're spoilt these days!). This should be as easy as issuing:
docker pull jpebe/ari5902
You should copy all the Jupyter notebooks and data files here in a local directory and then issue:
docker run -ti --rm -v /your/local/dir/ARI5902_Research_Topics_In_Artificial_Intelligence/jupyter:/notebooks -p 8888:8888 jpebe/ari5902
You should then be able to copy the Jupyter notebook URL from the terminal into your browser (ctrl-click will open a browser automatically). We are now good to go!
The more docker-experienced people, can probably build the docker image from the docker file in this repository (docker build -t ari5902 .).
You may want to reclaim some space on your Ubuntu installation. Remove all images and containers in the following way.
docker rm -f $(docker ps -a -q)
docker rmi -f $(docker images -a -q)Pull requests (with fixes) will be recieved with thanks.
