CIBERSORTx

Overview

CIBERSORTx is a computational deconvolution algorithm that estimates the relative and absolute fractions of immune and stromal cell populations from bulk gene expression data (RNA-seq or microarray). It uses support vector regression (SVR) against a reference gene expression matrix (LM22 leukocyte signature matrix or custom signatures) to infer cell-type proportions without requiring physical cell sorting. The “X” version adds batch correction to reduce technology-dependent effects when applying a reference matrix across different profiling platforms. It is widely used for tumor microenvironment (TME) characterization in clinical cohorts where only bulk RNA-seq is available.

Used by

  • Applied to bulk RNA-seq from 176 metastatic urothelial carcinoma patients in UC-GENOME to estimate immune cell fractions; identified enrichment of plasma cells, memory B cells, and activated dendritic cells in luminal subtypes (LumP), and higher T cell inflamed and IFNG signature scores in Ba/Sq tumors PMID:36333289
  • CIBERSORT applied to RNA-seq data from 45 melanoma biopsies for immune cell deconvolution; CD8+ T cells and NK cells increased and M1 macrophages decreased on-therapy in responders PMID:29033130

Notes

  • CIBERSORTx requires a pre-defined reference signature matrix; results may be sensitive to choice of reference and to batch effects between training and test cohorts.
  • Imperfect in distinguishing closely related cell subtypes (e.g., CD4+ T cell subsets) without high-resolution reference signatures.
  • Total RNA-seq (as used in UC-GENOME) vs. capture-based RNA-seq may introduce systematic differences in deconvolution results.
  • Absolute mode estimates depend on a calibration step that can vary across platforms; relative mode (fraction) estimates are more comparable across studies.

Sources

  • PMID:36333289 — UC-GENOME metastatic UC study; CIBERSORTx immune deconvolution of bulk RNA-seq to characterize TME composition by molecular subtype.

This page was processed by crosslinker on 2026-05-06. - PMID:29033130

This page was processed by entity-page-writer on 2026-05-15.