Immune Infiltration Analysis Module

Purpose

This module is based on IOBR and other R packages to perform tumor microenvironment (TME) immune infiltration analysis from bulk RNA-Seq data, supporting multiple mainstream immune infiltration algorithms, and outputting standardized result tables and visualization charts.

Supported immune infiltration and stromal scoring algorithms:

• CIBERSORT: Immune cell infiltration analysis based on linear support vector regression

• EPIC: Immune cell infiltration analysis based on constrained least squares method

• MCPcounter: Immune cell infiltration analysis based on gene set enrichment analysis

• xCell: Immune cell infiltration analysis based on gene set variation analysis

• TIMER: Immune cell infiltration analysis based on deconvolution algorithm

• quanTIseq: Quantitative immune cell infiltration analysis

• ESTIMATE: Tumor purity assessment based on gene expression profiles

Input File Examples

• expression Expression matrix file: Each row represents a gene name, each column represents a sample name, values are corresponding expression levels, tab-separated

GeneID	Sample1	Sample2	Sample3
GENE1	1.234	2.345	3.456
GENE2	4.567	5.678	6.789

• clinical Clinical information file: Each row represents a sample name, each column represents a clinical feature, tab-separated

SampleID	tissue_type.samples	age	gender
Sample1	Tumor	45	Male
Sample2	Normal	50	Female
Sample3	Tumor	55	Male

Running Method

SDAS bulkValidate ​immuneScore --expression fpkm.txt --clinical clinical.txt --group_col tissue_type.samples --group_type discrete --output result_dir

Input Parameter Description

Parameter	Required	Default	Description
--expression	Yes		Expression matrix file path. Tab-separated, rows: gene IDs, columns: sample IDs, values: FPKM/TPM etc., cannot be raw counts, cannot be log
--clinical	Yes		Clinical information file path. Tab-separated, rows: sample IDs, columns: clinical features
--group_col	Yes		Grouping column name (must exist in clinical information file)
--output	Yes		Output directory path
--group_type	No	discrete	Grouping type: discrete/continuous, default discrete

• --group_type Grouping Type Description

  • Discrete grouping (discrete): Directly use categorical variables in clinical information for grouping, e.g.: Tumor vs Normal, Stage I vs Stage II vs Stage III

  • Continuous grouping (continuous): Divide continuous variables into three groups by quantiles: Low, Medium, High

  Quantiles: 0%, 30%, 70%, 100%

Output Results Display

Result File	Description
tme_combine.txt	Combined results of all immune infiltration methods
tme_heatmap.png/pdf	TME immune infiltration heatmap
StromaScore_heatmap.png/pdf	Stromal score heatmap
<method>_stacked_bar.png/pdf	Stacked bar chart of immune cell proportions for each method
<method>_grouped_boxplot.png/pdf	Grouped boxplot for each method

• Immune infiltration combined result table: tme_combine.txt Each row represents one sample, each column represents infiltration scores of different immune cell types. In addition to different immune cell infiltration scores, there is also stromal score information. StromalScore_estimate is the total amount of stromal cells, ImmuneScore_estimate is the total amount of immune cells, ESTIMATEScore_estimate is the non-tumor component, i.e., the total amount of stromal cells and immune cells. If these three columns have negative scores, it means there is almost no component of this type. The TumorPurity_estimate column is the proportion of tumor cells, distributed from 0 to 1.

  SampleID	Macrophages_M2_CIBERSORT	CD8_T_Cells_EPIC	...
  Sample1	0.123	0.456	...
  Sample2	0.234	0.567	...

• Immune infiltration heatmap: tme_heatmap.png/pdf Shows immune infiltration score heatmap of all samples, color indicates score level. Group is the grouping type set by --group_col. The figure shows comparison between cancer area and non-cancer area. Methods are results of 6 immune infiltration methods in this module. Through this figure, you can see which methods distinguish target cells or similar cell types, and also see if there are overall differences between cancer area and normal tissue.

• Stromal score heatmap: StromaScore_heatmap.png/pdf Shows stromal score heatmap of samples. Stromal scores are used to quantify the degree of stromal cell and immune cell infiltration in tumor microenvironment (TME) and tumor purity.

• Immune infiltration method visualization diagrams (immune_plots directory)

1. Stacked bar chart <method>_stacked_bar.png/pdf: Proportions of various immune cell types in different samples/groups.

1. Grouped boxplot <method>_grouped_boxplot.png/pdf: Distribution of various immune cell types in different groups.