annotate aurora_wgcna.Rmd @ 0:66ef158fa85c draft

Uploaded
author spficklin
date Fri, 22 Nov 2019 19:44:46 -0500
parents
children b14e4bf568b0
Ignore whitespace changes - Everywhere: Within whitespace: At end of lines:
rev   line source
0
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
1 ---
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
2 title: 'Aurora Galaxy WGCNA Tool: Gene Co-Expression Network Construction & Analysis'
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
3 output:
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
4 pdf_document:
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
5 number_sections: false
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
6 ---
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
7
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
8 ```{r setup, include=FALSE, warning=FALSE, message=FALSE}
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
9 knitr::opts_chunk$set(error = FALSE, echo = FALSE)
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
10 ```
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
11
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
12 ```{r}
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
13 # Make a directory for saving the figures.
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
14 dir.create('figures', showWarnings = FALSE)
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
15 ```
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
16
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
17 # Introduction
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
18 This report contains step-by-step results from use of the [Aurora Galaxy](https://github.com/statonlab/aurora-galaxy-tools) Weighted Gene Co-expression Network Analysis [WGCNA](https://bmcbioinformatics.biomedcentral.com/articles/10.1186/1471-2105-9-559) tool. This tool wraps the WGCNA R package into a ready-to-use Rmarkdown file. It performs module discovery and network construction using a dataset and optional trait data matrix provided.
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
19
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
20 If you provided trait data, a second report will be available with results comparing the trait values to the identified modules.
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
21
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
22 This report was generated on:
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
23 ```{r}
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
24 format(Sys.time(), "%a %b %d %X %Y")
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
25 ```
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
26
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
27
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
28 ## About the Input Data
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
29 ### Gene Expression Matrix (GEM)
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
30 The gene expression data is an *n* x *m* matrix where *n* rows are the genes, *m* columns are the samples and the elements represent gene expression levels (derived either from Microarray or RNA-Seq). The matrix was provided in a file meething these rules:
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
31 - Housed in a comma-separated (CSV) file.
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
32 - The rows represent the gene expression levels
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
33 - The first column of each row is the gene, transcript or probe name.
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
34 - The header contains only the sample names and therefore is one value less than the remaining rows of the file.
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
35
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
36 ### Trait/Phenotype Matrix
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
37 The trait/phenotype data is an *n* x *m* matrix where *n* is the samples and *m* are the features such as experimental condition, biosample properties, traits or phenotype values. The matrix is stored in a comma-separated (CSV) file and has a header.
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
38
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
39 ## Parameters provided by the user.
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
40 The following describes the input arguments provided to this tool:
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
41 ```{r}
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
42
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
43 if (!is.null(opt$height_cut)) {
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
44 print('The cut height for outlier removal of the sample dendrogram:')
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
45 print(opt$cut_height)
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
46 }
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
47
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
48 if (!is.null(opt$power)) {
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
49 print('The power to which the gene expression data is raised:')
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
50 print(opt$power)
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
51 }
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
52 print('The minimal size for a module:')
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
53 print(opt$min_cluster_size)
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
54
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
55 print('The block size for dividing the GEM to reduce memory requirements:')
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
56 print(opt$block_size)
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
57
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
58 print('The hard threshold when generating the graph file:')
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
59 print(opt$hard_threshold)
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
60
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
61 print('The character string used to identify missing values in the GEM:')
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
62 print(opt$missing_value1)
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
63
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
64 if (!is.null(opt$trait_data)) {
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
65 print('The column in the trait data that contains the sample name:')
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
66 print(opt$sname_col)
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
67
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
68 print('The character string used to identify missing values in the trait data:')
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
69 print(opt$missing_value2)
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
70
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
71 print('Columns in the trait data that should be treated as categorical:')
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
72 print(opt$one_hot_cols)
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
73
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
74 print('Columns in the trait data that should be ignored:')
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
75 print(opt$ignore_cols)
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
76 }
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
77 ```
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
78
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
79 ## If Errors Occur
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
80 Please note, that if any of the R code encountered problems, error messages will appear in the report below. If an error occurs anywhere in the report, results should be thrown out. Errors are usually caused by improperly formatted input data or improper input arguments. Use the following checklist to find and correct potential errors:
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
81
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
82 - Do the formats for the input datasets match the requirements listed above.
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
83 - Do the values set for missing values match the values in the input files, and is the missing value used consistently within the input files (i.e you don't have more than one such as 0.0 and 0, or NA and 0.0)
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
84 - If trait data was provided, check that the column specified for the sample name is correct.
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
85 - The block size should not exceed 10,000 and should not be lower than 1,000.
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
86 - Ensure that the sample names and all headers in the trait/phenotype data only contain alpha-numeric and underscore characters.
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
87
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
88
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
89 # Expression Data
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
90
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
91 The content below shows the first 10 rows and 6 columns of the Gene Expression Matrix (GEM) file that was provided.
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
92
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
93 ```{r}
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
94 gem = read.csv(opt$expression_data, header = TRUE, row.names = 1, na.strings = opt$missing_value1)
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
95 #table_data = head(gem, 100)
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
96 #datatable(table_data)
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
97 gem[1:10,1:6]
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
98 ```
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
99
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
100 ```{r}
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
101 gemt = as.data.frame(t(gem))
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
102 ```
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
103
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
104 The next step is to check the data for low quality samples or genes. These have too many missing values or consist of genes with zero-variance. Samples and genes are removed if they are low quality. The `goodSamplesGenes` function of WGCNA is used to identify such cases. The following cell indicates if WGCNA identified any low quality genes or samples, and these were removed.
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
105
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
106
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
107 ```{r}
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
108 gsg = goodSamplesGenes(gemt, verbose = 3)
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
109
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
110 if (!gsg$allOK) {
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
111 gemt = gemt[gsg$goodSamples, gsg$goodGenes]
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
112 } else {
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
113 print('all genes are OK!')
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
114 }
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
115 ```
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
116
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
117
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
118 Hierarchical clustering can be used to explore the similarity of expression across the samples of the GEM. The following dendrogram shows the results of that clustering. Outliers typically appear on their own in the dendrogram. If a height was not specified to trim outlier samples, then the `cutreeDynamic` function is used to automatically find outliers, and then they are removed. If you do not approve of the automatically detected height, you can re-run this tool with a desired cut height. The two plots below show the dendrogram before and after outlier removal.
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
119
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
120 ```{r fig.align='center'}
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
121 sampleTree = hclust(dist(gemt), method = "average");
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
122
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
123 plotSampleDendro <- function() {
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
124 plot(sampleTree, main = "Sample Clustering Prior to Outlier Removal", sub="", xlab="",
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
125 cex.axis = 1, cex.main = 1, cex = 0.5)
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
126 }
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
127 png('figures/01-sample_dendrogram.png', width=6 ,height=5, units="in", res=300)
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
128 plotSampleDendro()
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
129 invisible(dev.off())
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
130 plotSampleDendro()
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
131 ```
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
132
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
133 ```{r}
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
134 if (is.null(opt$height_cut)) {
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
135 print("You did not specify a height for cutting the dendrogram. The cutreeDynamic function was used.")
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
136 clust = cutreeDynamic(sampleTree, method="tree", minClusterSize = opt$min_cluster_size)
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
137 keepSamples = (clust!=0)
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
138 gemt = gemt[keepSamples, ]
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
139 } else {
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
140 print("You specified a height for cutting of", opt$height_cut, ". The cutreeStatic function was used.")
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
141 clust = cutreeStatic(sampleTree, cutHeight = opt$height_cut, minSize = opt$min_cluster_size)
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
142 keepSamples = (clust==1)
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
143 gemt = gemt[keepSamples, ]
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
144 }
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
145 n_genes = ncol(gemt)
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
146 n_samples = nrow(gemt)
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
147 removed = length(which(keepSamples == FALSE))
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
148 if (removed == 1) {
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
149 print(paste("A total of", removed, "sample was removed"))
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
150 } else {
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
151 print(paste("A total of", removed, "samples were removed"))
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
152 }
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
153
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
154 # Write out the filtered GEM
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
155 write.csv(t(gemt), opt$filtered_GEM, quote=FALSE)
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
156 ```
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
157 A file named `filtered_GEM.csv` has been created. This file is a comma-separated file containing the original gene expression data but with outlier samples removed. If no outliers were detected this file will be identical to the original.
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
158
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
159 ```{r fig.align='center'}
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
160 sampleTree = hclust(dist(gemt), method = "average");
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
161
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
162 plotFilteredSampleDendro <- function() {
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
163 plot(sampleTree, main = "Sample Clustering After Outlier Removal", sub="", xlab="",
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
164 cex.axis = 1, cex.main = 1, cex = 0.5)
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
165 }
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
166 png('figures/02-filtered-sample_dendrogram.png', width=6 ,height=5, units="in", res=300)
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
167 plotFilteredSampleDendro()
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
168 invisible(dev.off())
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
169 plotFilteredSampleDendro()
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
170 ```
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
171
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
172 # Network Module Discovery
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
173
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
174 The first step in network module discovery is calculating similarity of gene expression. This is performed by comparing the expression of every gene with every other gene using a correlation test. However, the WGCNA authors suggest that raising the GEM to a power that best approximates scale-free behavior improves the quality of the final modules. However, the power to which the data should be raised is initially unknown. This is determined using the `pickSoftThreshold` function of WGCNA which iterates through a series of power values (usually between 1 to 20) and tests how well the data approximates scale-free behavior. The following table shows the results of those tests. The meaning of the table headers are:
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
175
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
176 - Power: The power tested
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
177 - SFT.R.sq: This is the scale free index, or the R.squared value of the undelrying regression model. It indicates how well the power-raised data appears scale free. The higher the value the more scale-free.
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
178 - slope: The slope of the regression line used to calculate SFT.R.sq
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
179 - trunacted.R.sq: The adjusted R.squared measure from the truncated exponential model used to calculate SFT.R.sq
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
180 - mean.k: The mean degree (degree is a measure of how connected a gene is to every other gene. The higher the number the more connected.)
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
181 - median.k: The median degree
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
182 - max.k: The largest degree.
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
183
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
184 ```{r}
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
185 powers = c(1:10, seq(12, 20, 2))
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
186 sft = pickSoftThreshold(gemt, powerVector = powers, verbose = 5)
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
187 ```
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
188
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
189 The following plots show how the scale-free index and mean connectivity change as the power is adjusted. The ideal power value for the network should be the value where there is a diminishing change in both the scale-free index and mean connectivity.
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
190
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
191 ```{r fig.align='center'}
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
192 par(mfrow=c(1,2))
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
193 th = sft$fitIndices$SFT.R.sq[which(sft$fitIndices$Power == sft$powerEstimate)]
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
194
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
195 plotPower <- function() {
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
196 # Scale-free topology fit index as a function of the soft-thresholding power.
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
197 plot(sft$fitIndices[,1], -sign(sft$fitIndices[,3])*sft$fitIndices[,2],
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
198 xlab="Soft Threshold (power)",
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
199 ylab="Scale Free Topology Model Fit,signed R^2", type="n",
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
200 main = paste("Scale Independence"), cex.lab = 0.5);
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
201 text(sft$fitIndices[,1], -sign(sft$fitIndices[,3])*sft$fitIndices[,2],
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
202 labels=powers,cex=0.5,col="red");
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
203 #abline(h=th, col="blue")
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
204
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
205 # Mean connectivity as a function of the soft-thresholding power.
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
206 plot(sft$fitIndices[,1], sft$fitIndices[,5],
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
207 xlab="Soft Threshold (power)",ylab="Mean Connectivity", type="n",
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
208 main = paste("Mean Connectivity"), cex.lab = 0.5)
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
209 text(sft$fitIndices[,1], sft$fitIndices[,5], labels=powers, cex=0.5,col="red")
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
210 #abline(h=th, col="blue")
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
211 par(mfrow=c(1,1))
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
212 }
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
213
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
214 png('figures/03-power_thresholding.png', width=6 ,height=5, units="in", res=300)
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
215 plotPower()
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
216 invisible(dev.off())
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
217 plotPower()
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
218 ```
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
219 Using the values in the table above, WGCNA is able to predict the ideal power. This selection is indicated in the following cell and is shown as a blue line on the plots above. If you believe that the power was incorrectly chosen, you can re-run this tool with the same input files and provide the desired power.
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
220
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
221 ```{r}
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
222 print("WGCNA predicted the following power:")
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
223 print(sft$powerEstimate)
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
224 power = sft$powerEstimate
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
225 if (!is.null(opt$power)) {
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
226 print("However, you selected to override this by providing a power of:", opt$soft_threshold_power)
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
227 print(opt$soft_threshold_power)
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
228 power = opt$power
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
229 }
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
230 ```
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
231
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
232 Now that a power has been identified, modules can be discovered. Here, the `blockwiseModule` function of WGCNA is called. The dataset is divided into blocks of genes in order to keep memory usage low. The output of that function call is shown below. The number of blocks is dependent on the block size you provided.
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
233
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
234 ```{r}
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
235 net = blockwiseModules(gemt, power = power, maxBlockSize = opt$block_size,
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
236 TOMType = "unsigned", minModuleSize = opt$min_cluster_size,
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
237 reassignThreshold = 0, mergeCutHeight = 0.25,
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
238 numericLabels = TRUE, pamRespectsDendro = FALSE,
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
239 verbose = 1, saveTOMs = TRUE,
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
240 saveTOMFileBase = "TOM")
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
241 blocks = sort(unique(net$blocks))
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
242 ```
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
243 The following table shows the list of modules that were discovered and their size (i.e. number of genes).
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
244
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
245 ```{r}
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
246 module_labels = labels2colors(net$colors)
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
247 module_labels = paste("ME", module_labels, sep="")
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
248 module_labels2num = unique(data.frame(label = module_labels, num = net$color, row.names=NULL))
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
249 rownames(module_labels2num) = paste0('ME', module_labels2num$num)
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
250 modules = unique(as.data.frame(table(module_labels)))
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
251 n_modules = length(modules) - 1
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
252 module_size_upper = modules[2]
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
253 module_size_lower = modules[length(modules)]
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
254 colnames(modules) = c('Module', 'Module Size')
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
255 #datatable(modules)
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
256 modules
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
257 ```
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
258
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
259 Modules consist of a set of genes that have highly similar expression patterns. Therefore, the similarity of genes within a module can be summarized using an "eigengene" vector. This vector is analgous to the first principal component in a PCA analysis. Once each module's eigengene is calculated, they can be compared and displayed in dendrogram to identify which modules are most similar to each other. This is visible in the following plot.
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
260
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
261 ```{r fig.align='center'}
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
262 MEs = net$MEs
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
263 colnames(MEs) = module_labels2num[colnames(MEs),]$label
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
264
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
265 png('figures/04-module_dendrogram.png', width=6 ,height=5, units="in", res=300)
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
266 plotEigengeneNetworks(MEs, "Module Eigengene Dendrogram", plotHeatmaps = FALSE)
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
267 dev.off()
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
268 plotEigengeneNetworks(MEs, "Module Eigengene Dendrogram", plotHeatmaps = FALSE)
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
269 ```
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
270
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
271 Alternatively, we can use a heatmap to explore similarity of each module.
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
272 ```{r fig.align='center'}
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
273 plotModuleHeatmap <- function() {
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
274 plotEigengeneNetworks(MEs, "Module Eigengene Heatmap",
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
275 marHeatmap = c(2, 3, 2, 2),
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
276 plotDendrograms = FALSE)
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
277 }
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
278 png('figures/05-module_eigengene_heatmap.png', width=4 ,height=4, units="in", res=300)
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
279 plotModuleHeatmap()
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
280 invisible(dev.off())
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
281 plotModuleHeatmap()
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
282 ```
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
283
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
284 We can examine gene similarity within the context of our modules. The following dendrogram clusters genes by their similarity of expression and the modules to which each gene belongs is shown under the graph. When similar genes appear in the same module, the same colors will be visible in "blocks" under the dendrogram. The presence of blocks of color indicate that genes in modules tend to have similar expression.
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
285
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
286 ```{r}
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
287 # Plot the dendrogram and the module colors underneath
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
288 for (i in blocks) {
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
289 options(repr.plot.width=15, repr.plot.height=10)
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
290 colors = module_labels[net$blockGenes[[i]]]
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
291 colors = sub('ME','', colors)
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
292 plotClusterDendro <- function() {
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
293 plotDendroAndColors(net$dendrograms[[i]], colors,
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
294 "Module colors", dendroLabels = FALSE, hang = 0.03,
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
295 addGuide = TRUE, guideHang = 0.05,
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
296 main=paste('Cluster Dendgrogram, Block', i))
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
297 }
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
298 png(paste0('figures/06-cluster_dendrogram_block_', i, '.png'), width=6 ,height=4, units="in", res=300)
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
299 plotClusterDendro();
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
300 invisible(dev.off())
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
301 plotClusterDendro();
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
302 }
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
303
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
304 ```
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
305
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
306 The network is housed in a *n* x *n* similarity matrix known as the the Topological Overlap Matrix (TOM), where *n* is the number of genes and the value in each cell indicates the measure of similarity in terms of correlation of expression and interconnectedness. The following heat maps shows the TOM. Note, that the dendrograms in the TOM heat map may differ from what is shown above. This is because a subset of genes were selected to draw the heat maps in order to save on computational time.
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
307
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
308 ```{r}
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
309 for (i in blocks) {
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
310 # Load the TOM from a file.
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
311 load(net$TOMFiles[i])
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
312 TOM_size = length(which(net$blocks == i))
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
313 TOM = as.matrix(TOM, nrow=TOM_size, ncol=TOM_size)
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
314 dissTOM = 1-TOM
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
315
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
316 # For reproducibility, we set the random seed
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
317 set.seed(10);
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
318 select = sample(dim(TOM)[1], size = 1000);
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
319 selectColors = module_labels[net$blockGenes[[i]][select]]
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
320 selectTOM = dissTOM[select, select];
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
321
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
322 # There’s no simple way of restricting a clustering tree to a subset of genes, so we must re-cluster.
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
323 selectTree = hclust(as.dist(selectTOM), method = "average")
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
324
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
325 # Taking the dissimilarity to a power, say 10, makes the plot more informative by effectively changing
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
326 # the color palette; setting the diagonal to NA also improves the clarity of the plot
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
327 plotDiss = selectTOM^7;
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
328 diag(plotDiss) = NA;
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
329 colors = sub('ME','', selectColors)
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
330
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
331 png(paste0('figures/06-TOM_heatmap_block_', i, '.png'), width=6 ,height=6, units="in", res=300)
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
332 TOMplot(plotDiss, selectTree, colors, main = paste('TOM Heatmap, Block', i))
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
333 dev.off()
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
334 TOMplot(plotDiss, selectTree, colors, main = paste('TOM Heatmap, Block', i))
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
335 }
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
336 ```
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
337
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
338 ```{r}
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
339 output = cbind(colnames(gemt), module_labels)
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
340 colnames(output) = c('Gene', 'Module')
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
341 write.csv(output, file = opt$gene_module_file, quote=FALSE, row.names=FALSE)
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
342 ```
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
343
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
344 A file has been generated named `gene_module_file.csv` which contains the list of genes and the modules they belong to.
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
345
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
346 The TOM serves as both a simialrity matrix and an adjacency matrix. The adjacency matrix is typically identical to a similarity matrix but with values above a set threshold set to 1 and values below set to 0. This is known as hard thresholding. However, WGCNA does not set values above a threshold to zero but leaves the values as they are, hence the word "weighted" in the WGCNA name. Additionally, it does not use a threshold at all, so no elements are set to 0. This approach is called "soft thresholding", because the pairwise weights of all genes contributed to discover of modules. The name "soft thresholding" may be a misnomer, however, because no thresholding in the traditional sense actually occurs.
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
347
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
348 Unfortunately, this "soft thresholding" approach can make creation of a graph representation of the network difficult. If we exported the TOM as a connected graph it would result in a fully complete graph and would be difficult to interpret. Therefore, we must still set a hard-threshold if we want to visualize connectivity in graph form. Setting a hard threshold, if too high can result in genes being excluded from the graph and a threshold that is too low can result in too many false edges in the graph.
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
349
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
350 ```{r}
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
351 edges = data.frame(fromNode= c(), toNode=c(), weight=c(), direction=c(), fromAltName=c(), toAltName=c())
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
352 for (i in blocks) {
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
353 # Load the TOM from a file.
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
354 load(net$TOMFiles[i])
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
355 TOM_size = length(which(net$blocks == i))
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
356 TOM = as.matrix(TOM, nrow=TOM_size, ncol=TOM_size)
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
357 colnames(TOM) = colnames(gemt)[net$blockGenes[[i]]]
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
358 row.names(TOM) = colnames(gemt)[net$blockGenes[[i]]]
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
359
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
360 cydata = exportNetworkToCytoscape(TOM, threshold = opt$hard_threshold)
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
361 edges = rbind(edges, cydata$edgeData)
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
362 }
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
363
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
364 edges$Interaction = 'co'
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
365 output = edges[,c('fromNode','toNode','Interaction', 'weight')]
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
366 colnames(output) = c('Source', 'Target', 'Interaction', 'Weight')
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
367 write.table(output, file = opt$network_edges_file, quote=FALSE, row.names=FALSE, sep="\t")
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
368 ```
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
369
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
370 Using the hard threshold parameter provided, a file has been generated named `network_edges.txt` which contains the list of edges. You can import this file into [Cytoscape](https://cytoscape.org/) for visualization. If you would like a larger graph, you must re-run the tool with a smaller threshold.
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
371
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
372 ```{r}
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
373 # Save this image for the next step which is optional if theuser
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
374 # provides a trait file.
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
375 save.image(file=opt$r_data)
66ef158fa85c Uploaded
spficklin
parents:
diff changeset
376 ```