Meeting 1 - GeorgeIniatis/Blood_Brain_Barrier_Drug_Prediction GitHub Wiki

Meeting Minutes

The supervisor gave a detailed introduction of the project and outlined a provisional timetable

  1. Literature review. The supervisor mentioned that he will send emails with helpful information in the following days.
  2. Data generation. Create a new dataset of drugs that are known to pass the barrier and of those that cannot. Could augment already made dataset from the literature review. Could also use automated google searches. The more data the better for our model in the future
  3. Feature engineering/extraction. Usage of QSAR
  4. Training the model. The model could produce a binary choice or a confidence %. Python libraries already exist and could be useful
  5. Evaluating the model
  6. At the very least produce a Jupyter notebook.
  7. (Optional/If there is time) Produce a web app.
  8. Writing dissertation

The supervisor suggested to use the first few weeks as efficiently as possible before any coursework is set

The supervisor and I agreed that Meeting Minutes will be provided as well as a Plan of Action for the following week after every meeting. These will be stored in the project’s wiki page at GitHub

A provisional weekly meeting was set for 2pm every Friday

Action Plan

  • Look at literature
  • Try to find some already existing datasets of drugs and their relation to the BBB

Update

  • Meeting was set for 1pm every Thursday over Zoom