Related Experiment Video
Updated: May 24, 2025
Evaluation of Tumor-infiltrating Leukocyte Subsets in a Subcutaneous Tumor Model
Published on: April 13, 2015
Identifying perturbations that boost T-cell infiltration into tumours via counterfactual learning of their spatial
Zitong Jerry Wang1, Abdullah S Farooq2, Yu-Jen Chen2
1Division of Biology and Biological Engineering, California Institute of Technology, Pasadena, CA, USA. jerry@westlake.edu.cn.
Abstract:
Cancer progression can be slowed down or halted via the activation of either endogenous or engineered T cells and their infiltration of the tumour microenvironment. Here we describe a deep-learning model that uses large-scale spatial proteomic profiles of tumours to generate minimal tumour perturbations that boost T-cell infiltration. The model integrates a counterfactual optimization strategy for the generation of the perturbations with the prediction of T-cell infiltration as a self-supervised machine learning problem. We applied the model to 368 samples of metastatic melanoma and colorectal cancer assayed using 40-plex imaging mass cytometry, and discovered cohort-dependent combinatorial perturbations (CXCL9, CXCL10, CCL22 and CCL18 for melanoma, and CXCR4, PD-1, PD-L1 and CYR61 for colorectal cancer) that support T-cell infiltration across patient cohorts, as confirmed via in vitro experiments. Leveraging counterfactual-based predictions of spatial omics data may aid the design of cancer therapeutics.
Insights
A new deep-learning model predicts minimal tumor changes to enhance T-cell infiltration, a crucial step in cancer immunotherapy. This approach identifies specific molecular combinations to boost anti-cancer immune responses in melanoma and colorectal cancer.
Area of Science:
- Computational biology
- Immunology
- Oncology
Background:
- T-cell infiltration into the tumor microenvironment is critical for halting cancer progression.
- Current strategies often lack precision in modulating the tumor microenvironment to optimize T-cell activity.
Purpose of the Study:
- To develop a deep-learning model for predicting minimal tumor perturbations that enhance T-cell infiltration.
- To identify specific molecular combinations that promote T-cell infiltration in different cancer types.
Main Methods:
- Utilized a deep-learning model integrating counterfactual optimization and self-supervised learning for T-cell infiltration prediction.
- Applied the model to spatial proteomic profiles from 368 metastatic melanoma and colorectal cancer samples using 40-plex imaging mass cytometry.
- Validated predicted perturbations through in vitro experiments.
Main Results:
- Discovered cohort-dependent combinatorial perturbations that significantly boost T-cell infiltration.
- Identified specific molecular combinations for melanoma (CXCL9, CXCL10, CCL22, CCL18) and colorectal cancer (CXCR4, PD-1, PD-L1, CYR61).
- Confirmed the efficacy of these perturbations in supporting T-cell infiltration across patient cohorts.
Conclusions:
- Counterfactual-based predictions from spatial omics data can guide the design of novel cancer therapeutics.
- This deep-learning approach offers a promising strategy for enhancing immunotherapy by optimizing T-cell infiltration.
- The identified perturbations represent potential targets for developing more effective cancer treatments.
Related Concept Videos
01:27Tumor Immunotherapy
01:22T Cell Activation and Clonal Selection
Naive T cells that have not yet encountered an antigen express two primary CD...