Create a Pipeline from data extraction to execution of GNN based Drug Response Prediction

The use of machine learning models in medicine is expanding, but there are still undergoing due to the interpretability. One useful example which has understandability is DrugCell, which reflects biological structures in machine learning models. Based on this model, this project aims to build a machine learning pipeline using DrugCell from the data in the CellMinerCDB database developed by NRNB.

The current data structure of CellMinerCDB makes it challenging to pass inputs directly to DrugCell. Therefore, in this project, we will make PRs for 1) processing the data, 2) passing the data, and 3) calling DrugCell to make predictions.

The deliverable of this project is to merge the above PRs into rcellminer, so that data acquisition, processing, and prediction can be completed by rcellminer.