Hi HN,I ve been working on QUBE Predict, a cloud platform for drug response prediction built around graph-based biological descriptor learning.Rather than relying solely on end-to-end models, the framework focuses on reusable biological descriptor representations that can be paired with conventional machine learning algorithms.The platform includes:• REST API• Web interface• Batch prediction• Authentication and API keys• Cloud deploymentThe methodology has been described in a publicly available technical paper.Paper:https://doi.org/10.5281/zenodo.21455765Live Demo:https://qube-predict.streamlit.app/GitHub:https://github.com/gulfamh1999-afk/qube-predict-uiI d really appreciate technical feedback, especially on the representation-learning methodology and platform architecture.