Sensitive detection of synthetic response to cancer immunotherapy driven by gene paralog pairs

Area of Science:

• Oncology
• Genetics
• Immunology

Background:

• Cancer immunotherapies like checkpoint blockade and CAR-T therapy have transformed treatment but face challenges with patient response and relapse.
• Identifying single gene targets to improve immunotherapy is established, but finding effective combinations of genes remains difficult.

Purpose of the Study:

• To develop a computational approach for identifying cancer-intrinsic paralog pairs that can enhance immunotherapy efficacy.
• To create a predictive model for gene pairs influencing immunotherapy response.
• To experimentally validate the identified paralog pairs.

Main Methods:

• Utilized sgRNA set enrichment analysis on genome-wide screens to identify potential paralog pairs.
• Developed an ensemble learning model with an XGBoost classifier to predict paralog gene pairs impacting immunotherapy.
• Performed CRISPR double knockout experiments to validate the functional significance of predicted paralog pairs.

Main Results:

• Successfully identified cancer-intrinsic paralog pairs with the potential to enhance immunotherapy.
• The predictive model demonstrated effectiveness in pinpointing influential gene pairs.
• Experimental validation confirmed the functional role of these paralog pairs in immunotherapy response.

Conclusions:

• The developed computational approach offers a sensitive method for discovering novel paralog gene pairs that significantly impact cancer immunotherapy.
• This strategy can identify synergistic targets that improve treatment outcomes, even when individual genes show limited effects.