使用CEMiTool进行共表达模块分析

发表于 更新于 分类于 Valine：

使用 CEMiTools 进行共表达模块分析

CMEiTools 链接：https://cemitool.sysbio.tools/

CEMiTool 是一款基于 WGCNA 但又有别于 WGCNA 的一个 R 软件包，可以实现共表达模块的自动识别。

1. 数据文件准备

1.1 表达量文件

1
2
3
4
5
6
7
           sample1   sample2   sample3   sample4
gene1    13.061894 13.290272 13.360468 13.178729
gene2     3.410819  3.164874  3.599792  3.400613
gene3     6.326861  5.915121  6.341564  5.905167
gene4     3.237749  3.362508  3.320674  3.365530
gene5     3.980988  4.201731  4.235020  4.046716
gene6     3.379857  3.229973  3.150274  3.196610

1.2 样本分组信息

1
2
3
4
5
6
7
	SampleName Class
1   sample1    g0
2   sample2    g0
3   sample3    g0
4   sample4    g1
5   sample5    g1
6   sample6    g1

1.3 基因注释文件

基因注释文件需要提供 gmt 格式文件。

1.4 基因互作信息文件

1
2
3
4
5
6
7
  gene1symbol gene2symbol
1         DBH      REPIN1
2      RBFOX2       HERC5
3      ZNF460      CCDC22
4     SNRNP40        OAZ3
5       SRSF6        OAZ3
6      SPTAN1       ARL8A

2. 在线软件使用

将准备好的文件上传至 https://cemitool.sysbio.tools/analysis 网站进行在线分析。

3. 软件本地使用

CMEiTool 软件是一个 R 包，所以使用 R 语言安装载入后即可使用。

1
2
3
4
5
> if (!requireNamespace("BiocManager", quietly = TRUE))
    install.packages("BiocManager")

> BiocManager::install("CEMiTool")
> library("CEMiTool")

软件包下载地址：https://www.bioconductor.org/packages/release/bioc/html/CEMiTool.html

具体使用点击查看: 参考链接

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
 library(CEMiTool)

##
setwd("example/")
rm(list = ls())
##
expr <- read.table(file = "cemitool-expression.tsv",header = T, sep = "\t", row.names = 1)
sample_annot <- read.delim(file = "cemitool-phenotypes.tsv", header = T, sep = "\t")
gmt_in <- read_gmt(fname = "cemitool-genesets.gmt")

int_df <- read.delim(file = "cemitool-interactions.tsv", header = T, sep = "\t")

cem <- cemitool(expr=expr, 
                annot=sample_annot,
                gmt=gmt_in,
                interactions=int_df, 
                filter_pval = 0.05, 
                min_ngen = 30, 
                apply_vst = TRUE, 
                verbose=TRUE, 
                filter = TRUE, 
                cor_method = "pearson",
                network_type = "unsigned", 
                tom_type = "unsigned", 
                set_beta = NULL, 
                plot=TRUE
              )
# Create report as html file
generate_report(cem, directory = "Report", output_format="html_document")
# Write analysis results into files
write_files(cem, directory="Tables", force=TRUE)
# Save all plots
save_plots(cem, "all", directory="Plots")

4. Docker 容器中使用 CEMiTool 软件

首先下载安装 docker，然后使用 docker pull csblusp/cemitool 命令将 CEMiTool 软件拖至本地。

查看使用说明：

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
➜  ~ docker run csblusp/cemitool -h
CEMiTool - Co-Expression Modules identification Tool

Usage: cemitool.R EXPRSFILE  --output=<dir> [--sample-annot=<annot> --samples-column=<samplecol> --class-column=<classcol> --no-filter (--filter-pval=<p>|--ngenes=<ngenes>) --vst --eps --network-type=<nettype> --tom-type=<tomtype> --interactions=<inter> --pathways=<gmt> --ora-pvalue=<p> --cor-method=<cor> --no-merge --rank-method --min-module-size=<min> --diss-thresh=<thresh> --center-func=<fun> --directed --verbose]

Input:
  EXPRSFILE                         a normalized expression file .tsv format

Options:
  -h --help                          show this help message
  --version                          show program version
  -s <annot> --sample-annot=<annot>  sample annotation, must have a column with sample names and class
  --samples-column=<samplecol>       the column name containing sample names in template file [default: SampleName]
  --class-column=<classcol>          the column name containing classes in template file [default: Class]
  -i <int> --interactions=<int>      gene interaction file, must have two columns
  -p <gmt> --pathways=<gmt>          GMT file name (Gene Matrix Transposed format)
  --ora-pvalue=<p>                   p-value cutoff to be used on over representation analysis [default: 0.05]
  --no-filter                        does not filter the expression data.frame
  --filter-pval=<p>                  p-value to be used in the filtering step [default: 0.1]
  --vst                              apply Variance Stabilizing Transformation
  --ngenes=<ngenes>                  number of genes remaining after filtering
  --eps=<eps>                        epsilon [default: 0.1]
  -c <cor> --cor-method=<cor>        correlation method (spearman or pearson) [default: pearson]
  --network-type=<nettype>           network type, 'signed' or 'unsigned' [default: unsigned]
  --tom-type=<nettype>               TOM type, 'signed' or 'unsigned' [default: signed]
  --no-merge                         does not merge related modules based on eigengene similarity
  --rank-method                      rank method [default: mean]
  --min-module-size=<min>            minimum module size [default: 30]
  --diss-thresh=<thresh>             module merging correlation threshold for eigengene similarity [default: 0.8]
  --center-func=<fun>                metric used for centering [default: mean]
  --directed                         the interactions are directed
  -o <dir> --output=<dir>            output directory
  -v --verbose                       display progress messages

Authors:
  Pedro S T Russo - pedro.russo at usp.br
  Gustavo R Ferreira - gustavo.rodrigues.ferreira at usp.br
  Lucas E Cardozo - lucasecardozo at usp.br
  Matheus C Burger - burger at usp.br
  Thiago D C Hirata - thiagodch at gmail.com
  Diogenes S Lima - diogenes.lima at usp.br
  Fernando M Passos - fmarcon at usp.br
  Raul A Carrasco
  Melissa Lever - melissalever at gmail.com
  Vinicius Maracaja-Coutinho
  Helder I Nakaya - hnakaya at usp.br

More information:
  www.csbiology.com
  Computational Systems Biology Laboratory
  University of Sao Paulo, Brazil
Alipe wechat
扫码关注小刘哥订阅号！
谢谢您的赏识！