Abstract:Tumor microenvironment (TME) plays a key role in tumorigenesis, disease progression and the acquirement of drug resistance. An accurate assessment of TME cellular compositions may not only shed light on how TME interacts with cancer cells, but also bring new insights for translational researchers in studying the non-responding mechanism of immunotherapy. Traditional deconvolution methods infer the relative proportions of predefined cell types based on a tissue omics data through either regression- or enrichment- based approaches. However, there are several challenges that remain unsolved in the current formulations, including (1) identifying the Immune/Stromal (I/S) cell types that truly exists in a TME, (2) identifying the marker genes for each cell type that are specifically expressed by one or a few I/S cell types in a TME, (3) co-linearity among to-be-assessed I/S proportions due to their co-infiltration. We have developed a novel semi-supervised deconvolution method namely ICTD (Inference of Cell types and Deconvolution), addressing the three challenges via (i) developing a Bi-Cross Validation (BCV) based matrix rank test to assess the significance level of the existence of cell types and signature genes, (ii) utilizing a constrained Non-negative Matrix Factorization (NMF) to eliminate the effect of co-linearity. We validated ICTD on bulk tumor datasets simulated using single-cell RNA-seq data. Our analysis suggested that ICTD has a largely improved prediction accuracy of TME compositions comparing to existing methods, and particularly, it is capable of identifying novel or sub-cell types. We applied our method to TCGA and other gene expression data of breast and prostate cancer. Subsets of CD8+ T cells with varied cytotoxicity levels and subtypes of fibroblast cells were identified. Moreover, integrated with an analysis of an independent single-cell and cell line gene expression dataset, we identified genes specifically expressed by cancer cells that are associated with decreased T cell cytotoxicity level. Citation Format: Wennan Chang, Changlin Wan, Yifan Sun, Yan Han, Siyuan Qi, Xiongbin Lu, Sha Cao, Chi Zhang. A semi-supervised deconvolution method for quantifying the composition and activity of tumor-infiltrating cell types [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2019; 2019 Mar 29-Apr 3; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2019;79(13 Suppl):Abstract nr 2439.

ReCIDE: robust estimation of cell type proportions by integrating single-reference-based deconvolutions

Reassessment of Reliability and Reproducibility for Triple-Negative Breast Cancer Subtyping

Robust and accurate estimation of cellular fraction from tissue omics data via ensemble deconvolution

Adaptive Regularized Tri-Factor Non-Negative Matrix Factorization for Cell Type Deconvolution

Deconvolution Of Heterogeneous Tumor Samples Using Partial Reference Signals

ICTD: Inference of cell types and deconvolution -- a next-generation deconvolution method for accurate assess cell population and activities in tumor microenvironment.

Statistical Inference for Cell Type Deconvolution

A semi supervised approach for cell phenotypic and functional estimation in tissue microenvironment

DISSECT: deep semi-supervised consistency regularization for accurate cell type fraction and gene expression estimation

Abstract 2439: A Semi-Supervised Deconvolution Method for Quantifying the Composition and Activity of Tumor-Infiltrating Cell Types

DeepDecon Accurately Estimates Cancer Cell Fractions in Bulk RNA-seq Data

CATD: A reproducible pipeline for selecting cell-type deconvolution methods across tissues

BLEND: Probabilistic Cellular Deconvolution with Automated Reference Selection

Deconer: A comprehensive and systematic evaluation toolkit for reference-based cell type deconvolution algorithms using gene expression data

ICTD: A semi supervised cell type identification and deconvolution method for multi omics data

Spatial transcriptomics deconvolution at single-cell resolution using Redeconve

ARIC: accurate and robust inference of cell type proportions from bulk gene expression or DNA methylation data

Missing cell types in single-cell references impact deconvolution of bulk data but are detectable

imply: improving cell-type deconvolution accuracy using personalized reference profiles

Pathway-enhanced Transformer-based robust model for quantifying cell types of origin of cell-free transcriptome

NNICE: a deep quantile neural network algorithm for expression deconvolution