An Artificial Intelligence–Driven Multimodal Algorithm Predicts Immunotherapy and Targeted Therapy Outcomes in Clear Cell Renal Cell Carcinoma

Introduction

This repository supports our study on predictive modeling in metastatic clear cell renal cell carcinoma (ccRCC). While tyrosine kinase inhibitors (TKIs) and immune checkpoint inhibitors (ICIs) have transformed treatment, challenges such as toxicity and eventual resistance remain.

Summary

• Cohorts: Integrated transcriptomic data from 16 independent cohorts (n = 4,143 patients)
• Discovery: Identified five harmonized immune tumor microenvironment (HiTME) subtypes
• Validation: Confirmed HiTME classification using multiplex immunofluorescence (mIF)
• Model: Developed a machine learning–based predictive framework combining:
  • Genomic alterations
  • Transcriptomic features
  • Tumor microenvironment (TME) patterns
• Application: Predicts response to ICI and TKI therapies
• Clinical relevance: Enables retrospective clinical validation and paves the way for prospective trials

---

Repository Structure

├── data/                                  # Preprocessed expression and annotation data
│   ├── ccrcc_tki_0.1.pickle               # Trained TKI model
│   └── ccrcc_io_0.3.pickle                # Trained IO model
├── IO_model.ipynb                         # Jupyter notebook for IO model analysis and visualization
├── TKI_model.ipynb                        # Jupyter notebook for TKI model analysis and visualization
├── portraits/                             # Helper functions and pipelines
├── requirements.txt                       # Python package dependencies
├── make_tme_environment.sh                # Shell script for environment setup
├── RCC.AI_TKI_IO.Hsieh Notice File.md     # NOTICE OF OPEN SOURCE LICENSES, TERMS AND CONDITIONS
├── License.md                             # Software License Agreement
└── README.md                              # Project summary and instructions

Setup

git clone https://github.com/BostonGene/RCC.AI_TKI_IO.Hsieh.git
cd RCC.AI_TKI_IO.Hsieh
bash make_tme_environment.sh

Citation

For more information visit BostonGene’s Scientific portal.