This vignette illustrates how to easily use the CITMIC package. Here, we present a network-based computational approach, CTMIC, estimate cell infiltration by integrating gene transcriptomic data and individualized cell crosstalk network.
This network-based method major parts:
Constructing network and performing randomization. In our method, Here we collected a total of 86 cell types and 4171 GO-BPs. Among them, the gene characterization sets for various cell types came from 12 sources, including published cell signature sets (Bindea et al. [1], Charoentong et al. [2], Danaher et al. [3], Davoli et al. [4], He et al. [5], Rooney et al. [6], Tirosh et al. [7]) and the TME cell estimation methods (MCPcounter [8], EPIC [9], ImmuCellAI [10], TIDE [11] and xCell [12]). The molecular function gene sets were downloaded from C5 Gene sets derived from the GO biological process ontology in the Molecular Signatures Database (MSigDB). We constructed a Cell-GO bipartite network for each sample. We defined an edge between a Cell and a GO term if they have at least a common gene, and give the weight of this edge that is calculated by the Jaccard index and Gene Expression Profile(GEP). Next, we proposed to construct a Cell crosstalk network for each sample. The information from the Cell-GO bipartite network was used to evaluate the extent of interaction between two Cells as the edge’s weight. We consider that the extent of interaction between two Cells will be stronger when there are more GO terms they target in common and the more related they are to these GO terms. Then, we used the eigenvector centrality measure to calculate how central each Cell is in this individualized Cell crosstalk network. Finally, cell infiltration in the patient dataset was obtained by normalizing the centrality scores of the cells.
This package provides the GetData_CITMIC function to
return example data set and environment variables, such as the gene
expression matrix and so on.
##Example calculate cell infiltration.
The function CITMIC is used to calculate the Cell
infiltration (lnScore). This function requires users to input the gene
expression matrix, the weighted specifies whether to create a
weighted graph for the cell crosstalk network, the base for the
standardized log base of data, the damping is the restart
probability of random walk and the cl.cores is a number of CPU
cores applied to this task. cell.type preset the relevant cell
type (e.g. if the solid tumor tissue does not contain ‘HSC’, it is
better to remove it when we preset it). We recommend using as input the
gene expression matrix normalized by log2(fpkm+1). An example data is
stored in the package.
The commands are as follows:
#Detect the Number of CPU Cores (Select the number of cl.cores to apply to CITMIC based on detectCores())
library(parallel)
library(igraph)##
## Attaching package: 'igraph'## The following objects are masked from 'package:stats':
##
## decompose, spectrum## The following object is masked from 'package:base':
##
## union## [1] 4# Obtain the example data
GEP<-GetData_CITMIC("GEP")
# Calculate the Cell infiltration(lnScore)
lnScore<-CITMIC(GEP,weighted = TRUE,base = 10,damping=0.90,cl.cores=1,cell.type=NULL)
head(lnScore)## TCGA.EE.A2GJ.06A TCGA.EE.A2GI.06A TCGA.WE.A8ZM.06A
## Activated B cell 0.4155478 0.5319714 0.1549675
## Activated CD4+ T cells 0.6105840 0.6376140 0.3149371
## Activated CD8+ T cells 0.3987288 0.4678214 0.1503589
## aDCs 0.9157299 0.9146414 0.8568350
## Adipocytes 0.2190414 0.3170802 0.4874105
## Angiogenesis 0.5762410 0.5273987 0.6974899
## TCGA.DA.A1IA.06A TCGA.D3.A51H.06A
## Activated B cell 0.1314579 0.7365269
## Activated CD4+ T cells 0.4463615 0.7212100
## Activated CD8+ T cells 0.2170825 0.5600082
## aDCs 0.9070054 0.9060217
## Adipocytes 0.3171054 0.2778043
## Angiogenesis 0.5255911 0.50432801. Bindea, G., et al., Spatiotemporal dynamics of intratumoral immune cells reveal the immune landscape in human cancer. Immunity, 2013. 39(4): p. 782-95. 2. Charoentong, P., et al., Pan-cancer Immunogenomic Analyses Reveal Genotype-Immunophenotype Relationships and Predictors of Response to Checkpoint Blockade. Cell Rep, 2017. 18(1): p. 248-262. 3. Danaher, P., et al., Gene expression markers of Tumor Infiltrating Leukocytes. J Immunother Cancer, 2017. 5: p. 18. 4. Davoli, T., et al., Tumor aneuploidy correlates with markers of immune evasion and with reduced response to immunotherapy. Science, 2017. 355(6322). 5. He, Y., et al., Classification of triple-negative breast cancers based on Immunogenomic profiling. J Exp Clin Cancer Res, 2018. 37(1): p. 327. 6. Rooney, M.S., et al., Molecular and genetic properties of tumors associated with local immune cytolytic activity. Cell, 2015. 160(1-2): p. 48-61. 7. Tirosh, I., et al., Dissecting the multicellular ecosystem of metastatic melanoma by single-cell RNA-seq. Science, 2016. 352(6282): p. 189-96. 8. Becht, E., et al., Estimating the population abundance of tissue-infiltrating immune and stromal cell populations using gene expression. Genome Biol, 2016. 17(1): p. 218. 9. Racle, J., et al., Simultaneous enumeration of cancer and immune cell types from bulk tumor gene expression data. Elife, 2017. 6. 10. Miao, Y.R., et al., ImmuCellAI: A Unique Method for Comprehensive T-Cell Subsets Abundance Prediction and its Application in Cancer Immunotherapy. Adv Sci (Weinh), 2020. 7(7): p. 1902880. 11. Jiang, P., et al., Signatures of T cell dysfunction and exclusion predict cancer immunotherapy response. Nat Med, 2018. 24(10): p. 1550-1558. 12. Aran, D., Z. Hu, and A.J. Butte, xCell: digitally portraying the tissue cellular heterogeneity landscape. Genome Biol, 2017. 18(1): p. 220.