annotate rgedgeRpaired_nocamera.xml @ 144:ab78085c7847 draft

Uploaded
author iuc
date Fri, 10 Apr 2015 22:46:53 -0400
parents 1435811cbf01
children
Ignore whitespace changes - Everywhere: Within whitespace: At end of lines:
rev   line source
143
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
1 <?xml version="1.0"?>
132
0de946608423 Uploaded
fubar
parents:
diff changeset
2 <tool id="rgdifferentialcount" name="Differential_Count" version="0.28">
0de946608423 Uploaded
fubar
parents:
diff changeset
3 <description>models using BioConductor packages</description>
0de946608423 Uploaded
fubar
parents:
diff changeset
4 <requirements>
143
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
5 <requirement type="package" version="3.1.2">R</requirement>
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
6 <requirement type="package" version="1.3.18">graphicsmagick</requirement>
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
7 <requirement type="package" version="9.10">ghostscript</requirement>
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
8 <requirement type="package" version="2.14">biocbasics</requirement>
132
0de946608423 Uploaded
fubar
parents:
diff changeset
9 </requirements>
0de946608423 Uploaded
fubar
parents:
diff changeset
10 <command interpreter="python">
0de946608423 Uploaded
fubar
parents:
diff changeset
11 rgToolFactory.py --script_path "$runme" --interpreter "Rscript" --tool_name "Differential_Counts"
0de946608423 Uploaded
fubar
parents:
diff changeset
12 --output_dir "$html_file.files_path" --output_html "$html_file" --make_HTML "yes"
0de946608423 Uploaded
fubar
parents:
diff changeset
13 </command>
0de946608423 Uploaded
fubar
parents:
diff changeset
14 <inputs>
143
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
15 <param name="input1" type="data" format="tabular" label="Select an input matrix - rows are contigs, columns are counts for each sample" help="Use the HTSeq based count matrix preparation tool to create these matrices from BAM/SAM files and a GTF file of genomic features"/>
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
16 <param name="title" type="text" value="Differential Counts" size="80" label="Title for job outputs" help="Supply a meaningful name here to remind you what the outputs contain">
132
0de946608423 Uploaded
fubar
parents:
diff changeset
17 <sanitizer invalid_char="">
143
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
18 <valid initial="string.letters,string.digits">
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
19 <add value="_"/>
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
20 </valid>
132
0de946608423 Uploaded
fubar
parents:
diff changeset
21 </sanitizer>
0de946608423 Uploaded
fubar
parents:
diff changeset
22 </param>
0de946608423 Uploaded
fubar
parents:
diff changeset
23 <param name="treatment_name" type="text" value="Treatment" size="50" label="Treatment Name"/>
143
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
24 <param name="Treat_cols" label="Select columns containing treatment." type="data_column" data_ref="input1" numerical="True" multiple="true" use_header_names="true" size="120" display="checkboxes" force_select="True">
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
25 <validator type="no_options" message="Please select at least one column."/>
132
0de946608423 Uploaded
fubar
parents:
diff changeset
26 </param>
0de946608423 Uploaded
fubar
parents:
diff changeset
27 <param name="control_name" type="text" value="Control" size="50" label="Control Name"/>
143
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
28 <param name="Control_cols" label="Select columns containing control." type="data_column" data_ref="input1" numerical="True" multiple="true" use_header_names="true" size="120" display="checkboxes" force_select="True">
132
0de946608423 Uploaded
fubar
parents:
diff changeset
29 </param>
143
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
30 <param name="subjectids" type="text" optional="true" size="120" value="" label="IF SUBJECTS NOT ALL INDEPENDENT! Enter comma separated strings to indicate sample labels for (eg) pairing - must be one for every column in input" help="Leave blank if no pairing, but eg if data from sample id A99 is in columns 2,4 and id C21 is in 3,5 then enter 'A99,C21,A99,C21'">
132
0de946608423 Uploaded
fubar
parents:
diff changeset
31 <sanitizer>
143
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
32 <valid initial="string.letters,string.digits">
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
33 <add value=","/>
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
34 </valid>
132
0de946608423 Uploaded
fubar
parents:
diff changeset
35 </sanitizer>
0de946608423 Uploaded
fubar
parents:
diff changeset
36 </param>
143
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
37 <param name="fQ" type="float" value="0.3" size="5" label="Non-differential contig count quantile threshold - zero to analyze all non-zero read count contigs" help="May be a good or a bad idea depending on the biology and the question. EG 0.3 = sparsest 30% of contigs with at least one read are removed before analysis"/>
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
38 <param name="useNDF" type="boolean" truevalue="T" falsevalue="F" checked="false" size="1" label="Non differential filter - remove contigs below a threshold (1 per million) for half or more samples" help="May be a good or a bad idea depending on the biology and the question. This was the old default. Quantile based is available as an alternative"/>
132
0de946608423 Uploaded
fubar
parents:
diff changeset
39 <conditional name="edgeR">
143
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
40 <param name="doedgeR" type="select" label="Run this model using edgeR" help="edgeR uses a negative binomial model and seems to be powerful, even with few replicates">
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
41 <option value="F">Do not run edgeR</option>
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
42 <option value="T" selected="true">Run edgeR</option>
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
43 </param>
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
44 <when value="T">
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
45 <param name="edgeR_priordf" type="integer" value="10" size="3" label="prior.df for tagwise dispersion - larger value = more squeezing of tag dispersions to common dispersion. Replaces prior.n and prior.df = prior.n * residual.df" help="10 = edgeR default. Use a larger value to 'smooth' small samples. See edgeR docs and note below"/>
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
46 <param name="edgeR_robust_method" type="select" value="20" size="3" label="Use robust dispersion method" help="Use ordinary, anscombe or deviance robust deviance estimates">
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
47 <option value="ordinary" selected="true">Use ordinary deviance estimates</option>
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
48 <option value="deviance">Use robust deviance estimates</option>
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
49 <option value="anscombe">use Anscombe robust deviance estimates</option>
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
50 </param>
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
51 </when>
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
52 <when value="F"/>
132
0de946608423 Uploaded
fubar
parents:
diff changeset
53 </conditional>
0de946608423 Uploaded
fubar
parents:
diff changeset
54 <conditional name="DESeq2">
143
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
55 <param name="doDESeq2" type="select" label="Run the same model with DESeq2 and compare findings" help="DESeq2 is an update to the DESeq package. It uses different assumptions and methods to edgeR">
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
56 <option value="F" selected="true">Do not run DESeq2</option>
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
57 <option value="T">Run DESeq2</option>
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
58 </param>
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
59 <when value="T">
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
60 <param name="DESeq_fitType" type="select">
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
61 <option value="parametric" selected="true">Parametric (default) fit for dispersions</option>
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
62 <option value="local">Local fit - this will automagically be used if parametric fit fails</option>
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
63 <option value="mean">Mean dispersion fit- use this if you really understand what you're doing - read the fine manual linked below in the documentation</option>
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
64 </param>
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
65 </when>
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
66 <when value="F"> </when>
132
0de946608423 Uploaded
fubar
parents:
diff changeset
67 </conditional>
143
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
68 <param name="doVoom" type="select" label="Run the same model with Voom/limma and compare findings" help="Voom uses counts per million and a precise transformation of variance so count data can be analysed using limma">
132
0de946608423 Uploaded
fubar
parents:
diff changeset
69 <option value="F" selected="true">Do not run VOOM</option>
0de946608423 Uploaded
fubar
parents:
diff changeset
70 <option value="T">Run VOOM</option>
143
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
71 </param>
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
72 <param name="fdrthresh" type="float" value="0.05" size="5" label="P value threshold for FDR filtering for amily wise error rate control" help="Conventional default value of 0.05 recommended"/>
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
73 <param name="fdrtype" type="select" label="FDR (Type II error) control method" help="Use fdr or bh typically to control for the number of tests in a reliable way">
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
74 <option value="fdr" selected="true">fdr</option>
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
75 <option value="BH">Benjamini Hochberg</option>
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
76 <option value="BY">Benjamini Yukateli</option>
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
77 <option value="bonferroni">Bonferroni</option>
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
78 <option value="hochberg">Hochberg</option>
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
79 <option value="holm">Holm</option>
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
80 <option value="hommel">Hommel</option>
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
81 <option value="none">no control for multiple tests</option>
132
0de946608423 Uploaded
fubar
parents:
diff changeset
82 </param>
0de946608423 Uploaded
fubar
parents:
diff changeset
83 </inputs>
0de946608423 Uploaded
fubar
parents:
diff changeset
84 <outputs>
0de946608423 Uploaded
fubar
parents:
diff changeset
85 <data format="tabular" name="out_edgeR" label="${title}_topTable_edgeR.xls">
143
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
86 <filter>edgeR['doedgeR'] == "T"</filter>
132
0de946608423 Uploaded
fubar
parents:
diff changeset
87 </data>
0de946608423 Uploaded
fubar
parents:
diff changeset
88 <data format="tabular" name="out_DESeq2" label="${title}_topTable_DESeq2.xls">
143
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
89 <filter>DESeq2['doDESeq2'] == "T"</filter>
132
0de946608423 Uploaded
fubar
parents:
diff changeset
90 </data>
0de946608423 Uploaded
fubar
parents:
diff changeset
91 <data format="tabular" name="out_VOOM" label="${title}_topTable_VOOM.xls">
143
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
92 <filter>doVoom == "T"</filter>
132
0de946608423 Uploaded
fubar
parents:
diff changeset
93 </data>
0de946608423 Uploaded
fubar
parents:
diff changeset
94 <data format="html" name="html_file" label="${title}.html"/>
0de946608423 Uploaded
fubar
parents:
diff changeset
95 </outputs>
143
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
96 <stdio>
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
97 <exit_code range="4" level="fatal" description="Number of subject ids must match total number of samples in the input matrix"/>
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
98 </stdio>
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
99 <tests>
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
100 <test>
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
101 <param name="input1" value="test_bams2mx.xls" ftype="tabular"/>
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
102 <param name="treatment_name" value="liver"/>
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
103 <param name="title" value="edgeRtest"/>
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
104 <param name="useNDF" value=""/>
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
105 <param name="doedgeR" value="T"/>
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
106 <param name="doVoom" value="T"/>
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
107 <param name="doDESeq2" value="T"/>
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
108 <param name="fdrtype" value="fdr"/>
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
109 <param name="edgeR_priordf" value="8"/>
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
110 <param name="edgeR_robust" value="ordinary"/>
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
111 <param name="fdrthresh" value="0.05"/>
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
112 <param name="control_name" value="heart"/>
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
113 <param name="subjectids" value=""/>
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
114 <param name="Control_cols" value="3,4,5,9"/>
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
115 <param name="Treat_cols" value="2,6,7,8"/>
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
116 <output name="out_edgeR" file="edgeRtest1out.xls" compare="diff" lines_diff="20"/>
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
117 <output name="html_file" file="edgeRtest1out.html" compare="diff" lines_diff="20"/>
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
118 </test>
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
119 </tests>
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
120 <configfiles>
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
121 <configfile name="runme"><![CDATA[
132
0de946608423 Uploaded
fubar
parents:
diff changeset
122 #
0de946608423 Uploaded
fubar
parents:
diff changeset
123 # edgeR.Rscript
0de946608423 Uploaded
fubar
parents:
diff changeset
124 # updated feb 2014 adding outlier-robust deviance estimate options by ross for R 3.0.2/bioc 2.13
0de946608423 Uploaded
fubar
parents:
diff changeset
125 # updated npv 2011 for R 2.14.0 and edgeR 2.4.0 by ross
0de946608423 Uploaded
fubar
parents:
diff changeset
126 # Performs DGE on a count table containing n replicates of two conditions
0de946608423 Uploaded
fubar
parents:
diff changeset
127 #
0de946608423 Uploaded
fubar
parents:
diff changeset
128 # Parameters
0de946608423 Uploaded
fubar
parents:
diff changeset
129 #
0de946608423 Uploaded
fubar
parents:
diff changeset
130 # 1 - Output Dir
0de946608423 Uploaded
fubar
parents:
diff changeset
131
0de946608423 Uploaded
fubar
parents:
diff changeset
132 # Original edgeR code by: S.Lunke and A.Kaspi
0de946608423 Uploaded
fubar
parents:
diff changeset
133 reallybig = log10(.Machine\$double.xmax)
0de946608423 Uploaded
fubar
parents:
diff changeset
134 reallysmall = log10(.Machine\$double.xmin)
0de946608423 Uploaded
fubar
parents:
diff changeset
135 library("gplots")
0de946608423 Uploaded
fubar
parents:
diff changeset
136 library("edgeR")
140
c50bdc1b7fa7 Uploaded
fubar
parents: 138
diff changeset
137 library('stringr')
132
0de946608423 Uploaded
fubar
parents:
diff changeset
138 hmap2 = function(cmat,nsamp=100,outpdfname='heatmap2.pdf', TName='Treatment',group=NA,myTitle='title goes here')
0de946608423 Uploaded
fubar
parents:
diff changeset
139 {
0de946608423 Uploaded
fubar
parents:
diff changeset
140 # Perform clustering for significant pvalues after controlling FWER
0de946608423 Uploaded
fubar
parents:
diff changeset
141 samples = colnames(cmat)
0de946608423 Uploaded
fubar
parents:
diff changeset
142 gu = unique(group)
0de946608423 Uploaded
fubar
parents:
diff changeset
143 gn = rownames(cmat)
0de946608423 Uploaded
fubar
parents:
diff changeset
144 if (length(gu) == 2) {
0de946608423 Uploaded
fubar
parents:
diff changeset
145 col.map = function(g) {if (g==gu[1]) "#FF0000" else "#0000FF"}
0de946608423 Uploaded
fubar
parents:
diff changeset
146 pcols = unlist(lapply(group,col.map))
0de946608423 Uploaded
fubar
parents:
diff changeset
147 } else {
0de946608423 Uploaded
fubar
parents:
diff changeset
148 colours = rainbow(length(gu),start=0,end=4/6)
0de946608423 Uploaded
fubar
parents:
diff changeset
149 pcols = colours[match(group,gu)] }
0de946608423 Uploaded
fubar
parents:
diff changeset
150 dm = cmat[(! is.na(gn)),]
0de946608423 Uploaded
fubar
parents:
diff changeset
151 # remove unlabelled hm rows
0de946608423 Uploaded
fubar
parents:
diff changeset
152 nprobes = nrow(dm)
0de946608423 Uploaded
fubar
parents:
diff changeset
153 # sub = paste('Showing',nprobes,'contigs ranked for evidence of differential abundance')
0de946608423 Uploaded
fubar
parents:
diff changeset
154 if (nprobes > nsamp) {
0de946608423 Uploaded
fubar
parents:
diff changeset
155 dm =dm[1:nsamp,]
0de946608423 Uploaded
fubar
parents:
diff changeset
156 #sub = paste('Showing',nsamp,'contigs ranked for evidence for differential abundance out of',nprobes,'total')
0de946608423 Uploaded
fubar
parents:
diff changeset
157 }
0de946608423 Uploaded
fubar
parents:
diff changeset
158 newcolnames = substr(colnames(dm),1,20)
0de946608423 Uploaded
fubar
parents:
diff changeset
159 colnames(dm) = newcolnames
0de946608423 Uploaded
fubar
parents:
diff changeset
160 pdf(outpdfname)
0de946608423 Uploaded
fubar
parents:
diff changeset
161 heatmap.2(dm,main=myTitle,ColSideColors=pcols,col=topo.colors(100),dendrogram="col",key=T,density.info='none',
0de946608423 Uploaded
fubar
parents:
diff changeset
162 Rowv=F,scale='row',trace='none',margins=c(8,8),cexRow=0.4,cexCol=0.5)
0de946608423 Uploaded
fubar
parents:
diff changeset
163 dev.off()
0de946608423 Uploaded
fubar
parents:
diff changeset
164 }
0de946608423 Uploaded
fubar
parents:
diff changeset
165
0de946608423 Uploaded
fubar
parents:
diff changeset
166 hmap = function(cmat,nmeans=4,outpdfname="heatMap.pdf",nsamp=250,TName='Treatment',group=NA,myTitle="Title goes here")
0de946608423 Uploaded
fubar
parents:
diff changeset
167 {
0de946608423 Uploaded
fubar
parents:
diff changeset
168 # for 2 groups only was
0de946608423 Uploaded
fubar
parents:
diff changeset
169 #col.map = function(g) {if (g==TName) "#FF0000" else "#0000FF"}
0de946608423 Uploaded
fubar
parents:
diff changeset
170 #pcols = unlist(lapply(group,col.map))
0de946608423 Uploaded
fubar
parents:
diff changeset
171 gu = unique(group)
0de946608423 Uploaded
fubar
parents:
diff changeset
172 colours = rainbow(length(gu),start=0.3,end=0.6)
0de946608423 Uploaded
fubar
parents:
diff changeset
173 pcols = colours[match(group,gu)]
0de946608423 Uploaded
fubar
parents:
diff changeset
174 nrows = nrow(cmat)
0de946608423 Uploaded
fubar
parents:
diff changeset
175 mtitle = paste(myTitle,'Heatmap: n contigs =',nrows)
0de946608423 Uploaded
fubar
parents:
diff changeset
176 if (nrows > nsamp) {
0de946608423 Uploaded
fubar
parents:
diff changeset
177 cmat = cmat[c(1:nsamp),]
0de946608423 Uploaded
fubar
parents:
diff changeset
178 mtitle = paste('Heatmap: Top ',nsamp,' DE contigs (of ',nrows,')',sep='')
0de946608423 Uploaded
fubar
parents:
diff changeset
179 }
0de946608423 Uploaded
fubar
parents:
diff changeset
180 newcolnames = substr(colnames(cmat),1,20)
0de946608423 Uploaded
fubar
parents:
diff changeset
181 colnames(cmat) = newcolnames
0de946608423 Uploaded
fubar
parents:
diff changeset
182 pdf(outpdfname)
0de946608423 Uploaded
fubar
parents:
diff changeset
183 heatmap(cmat,scale='row',main=mtitle,cexRow=0.3,cexCol=0.4,Rowv=NA,ColSideColors=pcols)
0de946608423 Uploaded
fubar
parents:
diff changeset
184 dev.off()
0de946608423 Uploaded
fubar
parents:
diff changeset
185 }
0de946608423 Uploaded
fubar
parents:
diff changeset
186
0de946608423 Uploaded
fubar
parents:
diff changeset
187 qqPlot = function(descr='qqplot',pvector, outpdf='qqplot.pdf',...)
0de946608423 Uploaded
fubar
parents:
diff changeset
188 # stolen from https://gist.github.com/703512
0de946608423 Uploaded
fubar
parents:
diff changeset
189 {
0de946608423 Uploaded
fubar
parents:
diff changeset
190 o = -log10(sort(pvector,decreasing=F))
0de946608423 Uploaded
fubar
parents:
diff changeset
191 e = -log10( 1:length(o)/length(o) )
0de946608423 Uploaded
fubar
parents:
diff changeset
192 o[o==-Inf] = reallysmall
0de946608423 Uploaded
fubar
parents:
diff changeset
193 o[o==Inf] = reallybig
0de946608423 Uploaded
fubar
parents:
diff changeset
194 maint = descr
0de946608423 Uploaded
fubar
parents:
diff changeset
195 pdf(outpdf)
0de946608423 Uploaded
fubar
parents:
diff changeset
196 plot(e,o,pch=19,cex=1, main=maint, ...,
0de946608423 Uploaded
fubar
parents:
diff changeset
197 xlab=expression(Expected~~-log[10](italic(p))),
0de946608423 Uploaded
fubar
parents:
diff changeset
198 ylab=expression(Observed~~-log[10](italic(p))),
0de946608423 Uploaded
fubar
parents:
diff changeset
199 xlim=c(0,max(e)), ylim=c(0,max(o)))
0de946608423 Uploaded
fubar
parents:
diff changeset
200 lines(e,e,col="red")
0de946608423 Uploaded
fubar
parents:
diff changeset
201 grid(col = "lightgray", lty = "dotted")
0de946608423 Uploaded
fubar
parents:
diff changeset
202 dev.off()
0de946608423 Uploaded
fubar
parents:
diff changeset
203 }
0de946608423 Uploaded
fubar
parents:
diff changeset
204
0de946608423 Uploaded
fubar
parents:
diff changeset
205 smearPlot = function(myDGEList,deTags, outSmear, outMain)
0de946608423 Uploaded
fubar
parents:
diff changeset
206 {
0de946608423 Uploaded
fubar
parents:
diff changeset
207 pdf(outSmear)
0de946608423 Uploaded
fubar
parents:
diff changeset
208 plotSmear(myDGEList,de.tags=deTags,main=outMain)
0de946608423 Uploaded
fubar
parents:
diff changeset
209 grid(col="lightgray", lty="dotted")
0de946608423 Uploaded
fubar
parents:
diff changeset
210 dev.off()
0de946608423 Uploaded
fubar
parents:
diff changeset
211 }
0de946608423 Uploaded
fubar
parents:
diff changeset
212
0de946608423 Uploaded
fubar
parents:
diff changeset
213 boxPlot = function(rawrs,cleanrs,maint,myTitle,pdfname)
0de946608423 Uploaded
fubar
parents:
diff changeset
214 {
0de946608423 Uploaded
fubar
parents:
diff changeset
215 nc = ncol(rawrs)
0de946608423 Uploaded
fubar
parents:
diff changeset
216 ##### for (i in c(1:nc)) {rawrs[(rawrs[,i] < 0),i] = NA}
0de946608423 Uploaded
fubar
parents:
diff changeset
217 fullnames = colnames(rawrs)
0de946608423 Uploaded
fubar
parents:
diff changeset
218 newcolnames = substr(colnames(rawrs),1,20)
0de946608423 Uploaded
fubar
parents:
diff changeset
219 colnames(rawrs) = newcolnames
0de946608423 Uploaded
fubar
parents:
diff changeset
220 newcolnames = substr(colnames(cleanrs),1,20)
0de946608423 Uploaded
fubar
parents:
diff changeset
221 colnames(cleanrs) = newcolnames
0de946608423 Uploaded
fubar
parents:
diff changeset
222 defpar = par(no.readonly=T)
0de946608423 Uploaded
fubar
parents:
diff changeset
223 print.noquote('@@@ Raw contig counts by sample:')
0de946608423 Uploaded
fubar
parents:
diff changeset
224 print.noquote(summary(rawrs))
0de946608423 Uploaded
fubar
parents:
diff changeset
225 print.noquote('@@@ Library size contig counts by sample:')
0de946608423 Uploaded
fubar
parents:
diff changeset
226 print.noquote(summary(cleanrs))
0de946608423 Uploaded
fubar
parents:
diff changeset
227 pdf(pdfname)
0de946608423 Uploaded
fubar
parents:
diff changeset
228 par(mfrow=c(1,2))
0de946608423 Uploaded
fubar
parents:
diff changeset
229 boxplot(rawrs,varwidth=T,notch=T,ylab='log contig count',col="maroon",las=3,cex.axis=0.35,main='log2 raw counts')
0de946608423 Uploaded
fubar
parents:
diff changeset
230 grid(col="lightgray",lty="dotted")
0de946608423 Uploaded
fubar
parents:
diff changeset
231 boxplot(cleanrs,varwidth=T,notch=T,ylab='log contig count',col="maroon",las=3,cex.axis=0.35,main=paste('log2 counts after ',maint))
0de946608423 Uploaded
fubar
parents:
diff changeset
232 grid(col="lightgray",lty="dotted")
0de946608423 Uploaded
fubar
parents:
diff changeset
233 dev.off()
0de946608423 Uploaded
fubar
parents:
diff changeset
234 pdfname = "sample_counts_histogram.pdf"
0de946608423 Uploaded
fubar
parents:
diff changeset
235 nc = ncol(rawrs)
0de946608423 Uploaded
fubar
parents:
diff changeset
236 print.noquote(paste('Using ncol rawrs=',nc))
0de946608423 Uploaded
fubar
parents:
diff changeset
237 ncroot = round(sqrt(nc))
0de946608423 Uploaded
fubar
parents:
diff changeset
238 if (ncroot*ncroot < nc) { ncroot = ncroot + 1 }
0de946608423 Uploaded
fubar
parents:
diff changeset
239 m = c()
0de946608423 Uploaded
fubar
parents:
diff changeset
240 for (i in c(1:nc)) {
0de946608423 Uploaded
fubar
parents:
diff changeset
241 rhist = hist(rawrs[,i],breaks=100,plot=F)
0de946608423 Uploaded
fubar
parents:
diff changeset
242 m = append(m,max(rhist\$counts))
0de946608423 Uploaded
fubar
parents:
diff changeset
243 }
0de946608423 Uploaded
fubar
parents:
diff changeset
244 ymax = max(m)
0de946608423 Uploaded
fubar
parents:
diff changeset
245 ncols = length(fullnames)
0de946608423 Uploaded
fubar
parents:
diff changeset
246 if (ncols > 20)
0de946608423 Uploaded
fubar
parents:
diff changeset
247 {
0de946608423 Uploaded
fubar
parents:
diff changeset
248 scale = 7*ncols/20
0de946608423 Uploaded
fubar
parents:
diff changeset
249 pdf(pdfname,width=scale,height=scale)
0de946608423 Uploaded
fubar
parents:
diff changeset
250 } else {
0de946608423 Uploaded
fubar
parents:
diff changeset
251 pdf(pdfname)
0de946608423 Uploaded
fubar
parents:
diff changeset
252 }
0de946608423 Uploaded
fubar
parents:
diff changeset
253 par(mfrow=c(ncroot,ncroot))
0de946608423 Uploaded
fubar
parents:
diff changeset
254 for (i in c(1:nc)) {
0de946608423 Uploaded
fubar
parents:
diff changeset
255 hist(rawrs[,i], main=paste("Contig logcount",i), xlab='log raw count', col="maroon",
0de946608423 Uploaded
fubar
parents:
diff changeset
256 breaks=100,sub=fullnames[i],cex=0.8,ylim=c(0,ymax))
0de946608423 Uploaded
fubar
parents:
diff changeset
257 }
0de946608423 Uploaded
fubar
parents:
diff changeset
258 dev.off()
0de946608423 Uploaded
fubar
parents:
diff changeset
259 par(defpar)
0de946608423 Uploaded
fubar
parents:
diff changeset
260
0de946608423 Uploaded
fubar
parents:
diff changeset
261 }
0de946608423 Uploaded
fubar
parents:
diff changeset
262
0de946608423 Uploaded
fubar
parents:
diff changeset
263 cumPlot = function(rawrs,cleanrs,maint,myTitle)
0de946608423 Uploaded
fubar
parents:
diff changeset
264 { # updated to use ecdf
0de946608423 Uploaded
fubar
parents:
diff changeset
265 pdfname = "Differential_rowsum_bar_charts.pdf"
0de946608423 Uploaded
fubar
parents:
diff changeset
266 defpar = par(no.readonly=T)
0de946608423 Uploaded
fubar
parents:
diff changeset
267 lrs = log(rawrs,10)
0de946608423 Uploaded
fubar
parents:
diff changeset
268 lim = max(lrs)
0de946608423 Uploaded
fubar
parents:
diff changeset
269 pdf(pdfname)
0de946608423 Uploaded
fubar
parents:
diff changeset
270 par(mfrow=c(2,1))
0de946608423 Uploaded
fubar
parents:
diff changeset
271 hist(lrs,breaks=100,main=paste('Before:',maint),xlab="# Reads (log)",
0de946608423 Uploaded
fubar
parents:
diff changeset
272 ylab="Count",col="maroon",sub=myTitle, xlim=c(0,lim),las=1)
0de946608423 Uploaded
fubar
parents:
diff changeset
273 grid(col="lightgray", lty="dotted")
0de946608423 Uploaded
fubar
parents:
diff changeset
274 lrs = log(cleanrs,10)
0de946608423 Uploaded
fubar
parents:
diff changeset
275 hist(lrs,breaks=100,main=paste('After:',maint),xlab="# Reads (log)",
0de946608423 Uploaded
fubar
parents:
diff changeset
276 ylab="Count",col="maroon",sub=myTitle,xlim=c(0,lim),las=1)
0de946608423 Uploaded
fubar
parents:
diff changeset
277 grid(col="lightgray", lty="dotted")
0de946608423 Uploaded
fubar
parents:
diff changeset
278 dev.off()
0de946608423 Uploaded
fubar
parents:
diff changeset
279 par(defpar)
0de946608423 Uploaded
fubar
parents:
diff changeset
280 }
0de946608423 Uploaded
fubar
parents:
diff changeset
281
0de946608423 Uploaded
fubar
parents:
diff changeset
282 cumPlot1 = function(rawrs,cleanrs,maint,myTitle)
0de946608423 Uploaded
fubar
parents:
diff changeset
283 { # updated to use ecdf
0de946608423 Uploaded
fubar
parents:
diff changeset
284 pdfname = paste(gsub(" ","", myTitle , fixed=TRUE),"RowsumCum.pdf",sep='_')
0de946608423 Uploaded
fubar
parents:
diff changeset
285 pdf(pdfname)
0de946608423 Uploaded
fubar
parents:
diff changeset
286 par(mfrow=c(2,1))
0de946608423 Uploaded
fubar
parents:
diff changeset
287 lastx = max(rawrs)
0de946608423 Uploaded
fubar
parents:
diff changeset
288 rawe = knots(ecdf(rawrs))
0de946608423 Uploaded
fubar
parents:
diff changeset
289 cleane = knots(ecdf(cleanrs))
0de946608423 Uploaded
fubar
parents:
diff changeset
290 cy = 1:length(cleane)/length(cleane)
0de946608423 Uploaded
fubar
parents:
diff changeset
291 ry = 1:length(rawe)/length(rawe)
0de946608423 Uploaded
fubar
parents:
diff changeset
292 plot(rawe,ry,type='l',main=paste('Before',maint),xlab="Log Contig Total Reads",
0de946608423 Uploaded
fubar
parents:
diff changeset
293 ylab="Cumulative proportion",col="maroon",log='x',xlim=c(1,lastx),sub=myTitle)
0de946608423 Uploaded
fubar
parents:
diff changeset
294 grid(col="blue")
0de946608423 Uploaded
fubar
parents:
diff changeset
295 plot(cleane,cy,type='l',main=paste('After',maint),xlab="Log Contig Total Reads",
0de946608423 Uploaded
fubar
parents:
diff changeset
296 ylab="Cumulative proportion",col="maroon",log='x',xlim=c(1,lastx),sub=myTitle)
0de946608423 Uploaded
fubar
parents:
diff changeset
297 grid(col="blue")
0de946608423 Uploaded
fubar
parents:
diff changeset
298 dev.off()
0de946608423 Uploaded
fubar
parents:
diff changeset
299 }
0de946608423 Uploaded
fubar
parents:
diff changeset
300
0de946608423 Uploaded
fubar
parents:
diff changeset
301
0de946608423 Uploaded
fubar
parents:
diff changeset
302
0de946608423 Uploaded
fubar
parents:
diff changeset
303 doGSEAold = function(y=NULL,design=NULL,histgmt="",
0de946608423 Uploaded
fubar
parents:
diff changeset
304 bigmt="/data/genomes/gsea/3.1/Abetterchoice_nocgp_c2_c3_c5_symbols_all.gmt",
0de946608423 Uploaded
fubar
parents:
diff changeset
305 ntest=0, myTitle="myTitle", outfname="GSEA.xls", minnin=5, maxnin=2000,fdrthresh=0.05,fdrtype="BH")
0de946608423 Uploaded
fubar
parents:
diff changeset
306 {
0de946608423 Uploaded
fubar
parents:
diff changeset
307 sink('Camera.log')
0de946608423 Uploaded
fubar
parents:
diff changeset
308 genesets = c()
0de946608423 Uploaded
fubar
parents:
diff changeset
309 if (bigmt > "")
0de946608423 Uploaded
fubar
parents:
diff changeset
310 {
0de946608423 Uploaded
fubar
parents:
diff changeset
311 bigenesets = readLines(bigmt)
0de946608423 Uploaded
fubar
parents:
diff changeset
312 genesets = bigenesets
0de946608423 Uploaded
fubar
parents:
diff changeset
313 }
0de946608423 Uploaded
fubar
parents:
diff changeset
314 if (histgmt > "")
0de946608423 Uploaded
fubar
parents:
diff changeset
315 {
0de946608423 Uploaded
fubar
parents:
diff changeset
316 hgenesets = readLines(histgmt)
0de946608423 Uploaded
fubar
parents:
diff changeset
317 if (bigmt > "") {
0de946608423 Uploaded
fubar
parents:
diff changeset
318 genesets = rbind(genesets,hgenesets)
0de946608423 Uploaded
fubar
parents:
diff changeset
319 } else {
0de946608423 Uploaded
fubar
parents:
diff changeset
320 genesets = hgenesets
0de946608423 Uploaded
fubar
parents:
diff changeset
321 } # use only history if no bi
0de946608423 Uploaded
fubar
parents:
diff changeset
322 }
0de946608423 Uploaded
fubar
parents:
diff changeset
323 print.noquote(paste("@@@read",length(genesets), 'genesets from',histgmt,bigmt))
0de946608423 Uploaded
fubar
parents:
diff changeset
324 genesets = strsplit(genesets,'\t') # tabular. genesetid\tURLorwhatever\tgene_1\t..\tgene_n
0de946608423 Uploaded
fubar
parents:
diff changeset
325 outf = outfname
0de946608423 Uploaded
fubar
parents:
diff changeset
326 head=paste(myTitle,'edgeR GSEA')
0de946608423 Uploaded
fubar
parents:
diff changeset
327 write(head,file=outfname,append=F)
0de946608423 Uploaded
fubar
parents:
diff changeset
328 ntest=length(genesets)
0de946608423 Uploaded
fubar
parents:
diff changeset
329 urownames = toupper(rownames(y))
0de946608423 Uploaded
fubar
parents:
diff changeset
330 upcam = c()
0de946608423 Uploaded
fubar
parents:
diff changeset
331 downcam = c()
0de946608423 Uploaded
fubar
parents:
diff changeset
332 for (i in 1:ntest) {
0de946608423 Uploaded
fubar
parents:
diff changeset
333 gs = unlist(genesets[i])
0de946608423 Uploaded
fubar
parents:
diff changeset
334 g = gs[1] # geneset_id
0de946608423 Uploaded
fubar
parents:
diff changeset
335 u = gs[2]
0de946608423 Uploaded
fubar
parents:
diff changeset
336 if (u > "") { u = paste("<a href=\'",u,"\'>",u,"</a>",sep="") }
0de946608423 Uploaded
fubar
parents:
diff changeset
337 glist = gs[3:length(gs)] # member gene symbols
0de946608423 Uploaded
fubar
parents:
diff changeset
338 glist = toupper(glist)
0de946608423 Uploaded
fubar
parents:
diff changeset
339 inglist = urownames %in% glist
0de946608423 Uploaded
fubar
parents:
diff changeset
340 nin = sum(inglist)
0de946608423 Uploaded
fubar
parents:
diff changeset
341 if ((nin > minnin) && (nin < maxnin)) {
0de946608423 Uploaded
fubar
parents:
diff changeset
342 ### print(paste('@@found',sum(inglist),'genes in glist'))
0de946608423 Uploaded
fubar
parents:
diff changeset
343 camres = camera(y=y,index=inglist,design=design)
0de946608423 Uploaded
fubar
parents:
diff changeset
344 if (! is.null(camres)) {
0de946608423 Uploaded
fubar
parents:
diff changeset
345 rownames(camres) = g # gene set name
0de946608423 Uploaded
fubar
parents:
diff changeset
346 camres = cbind(GeneSet=g,URL=u,camres)
0de946608423 Uploaded
fubar
parents:
diff changeset
347 if (camres\$Direction == "Up")
0de946608423 Uploaded
fubar
parents:
diff changeset
348 {
0de946608423 Uploaded
fubar
parents:
diff changeset
349 upcam = rbind(upcam,camres) } else {
0de946608423 Uploaded
fubar
parents:
diff changeset
350 downcam = rbind(downcam,camres)
0de946608423 Uploaded
fubar
parents:
diff changeset
351 }
0de946608423 Uploaded
fubar
parents:
diff changeset
352 }
0de946608423 Uploaded
fubar
parents:
diff changeset
353 }
0de946608423 Uploaded
fubar
parents:
diff changeset
354 }
0de946608423 Uploaded
fubar
parents:
diff changeset
355 uscam = upcam[order(upcam\$PValue),]
0de946608423 Uploaded
fubar
parents:
diff changeset
356 unadjp = uscam\$PValue
0de946608423 Uploaded
fubar
parents:
diff changeset
357 uscam\$adjPValue = p.adjust(unadjp,method=fdrtype)
0de946608423 Uploaded
fubar
parents:
diff changeset
358 nup = max(10,sum((uscam\$adjPValue < fdrthresh)))
0de946608423 Uploaded
fubar
parents:
diff changeset
359 dscam = downcam[order(downcam\$PValue),]
0de946608423 Uploaded
fubar
parents:
diff changeset
360 unadjp = dscam\$PValue
0de946608423 Uploaded
fubar
parents:
diff changeset
361 dscam\$adjPValue = p.adjust(unadjp,method=fdrtype)
0de946608423 Uploaded
fubar
parents:
diff changeset
362 ndown = max(10,sum((dscam\$adjPValue < fdrthresh)))
0de946608423 Uploaded
fubar
parents:
diff changeset
363 write.table(uscam,file=paste('camera_up',outfname,sep='_'),quote=F,sep='\t',row.names=F)
0de946608423 Uploaded
fubar
parents:
diff changeset
364 write.table(dscam,file=paste('camera_down',outfname,sep='_'),quote=F,sep='\t',row.names=F)
0de946608423 Uploaded
fubar
parents:
diff changeset
365 print.noquote(paste('@@@@@ Camera up top',nup,'gene sets:'))
0de946608423 Uploaded
fubar
parents:
diff changeset
366 write.table(head(uscam,nup),file="",quote=F,sep='\t',row.names=F)
0de946608423 Uploaded
fubar
parents:
diff changeset
367 print.noquote(paste('@@@@@ Camera down top',ndown,'gene sets:'))
0de946608423 Uploaded
fubar
parents:
diff changeset
368 write.table(head(dscam,ndown),file="",quote=F,sep='\t',row.names=F)
0de946608423 Uploaded
fubar
parents:
diff changeset
369 sink()
0de946608423 Uploaded
fubar
parents:
diff changeset
370 }
0de946608423 Uploaded
fubar
parents:
diff changeset
371
0de946608423 Uploaded
fubar
parents:
diff changeset
372
0de946608423 Uploaded
fubar
parents:
diff changeset
373
0de946608423 Uploaded
fubar
parents:
diff changeset
374
0de946608423 Uploaded
fubar
parents:
diff changeset
375 doGSEA = function(y=NULL,design=NULL,histgmt="",
0de946608423 Uploaded
fubar
parents:
diff changeset
376 bigmt="/data/genomes/gsea/3.1/Abetterchoice_nocgp_c2_c3_c5_symbols_all.gmt",
0de946608423 Uploaded
fubar
parents:
diff changeset
377 ntest=0, myTitle="myTitle", outfname="GSEA.xls", minnin=5, maxnin=2000,fdrthresh=0.05,fdrtype="BH")
0de946608423 Uploaded
fubar
parents:
diff changeset
378 {
0de946608423 Uploaded
fubar
parents:
diff changeset
379 sink('Camera.log')
0de946608423 Uploaded
fubar
parents:
diff changeset
380 genesets = c()
0de946608423 Uploaded
fubar
parents:
diff changeset
381 if (bigmt > "")
0de946608423 Uploaded
fubar
parents:
diff changeset
382 {
0de946608423 Uploaded
fubar
parents:
diff changeset
383 bigenesets = readLines(bigmt)
0de946608423 Uploaded
fubar
parents:
diff changeset
384 genesets = bigenesets
0de946608423 Uploaded
fubar
parents:
diff changeset
385 }
0de946608423 Uploaded
fubar
parents:
diff changeset
386 if (histgmt > "")
0de946608423 Uploaded
fubar
parents:
diff changeset
387 {
0de946608423 Uploaded
fubar
parents:
diff changeset
388 hgenesets = readLines(histgmt)
0de946608423 Uploaded
fubar
parents:
diff changeset
389 if (bigmt > "") {
0de946608423 Uploaded
fubar
parents:
diff changeset
390 genesets = rbind(genesets,hgenesets)
0de946608423 Uploaded
fubar
parents:
diff changeset
391 } else {
0de946608423 Uploaded
fubar
parents:
diff changeset
392 genesets = hgenesets
0de946608423 Uploaded
fubar
parents:
diff changeset
393 } # use only history if no bi
0de946608423 Uploaded
fubar
parents:
diff changeset
394 }
0de946608423 Uploaded
fubar
parents:
diff changeset
395 print.noquote(paste("@@@read",length(genesets), 'genesets from',histgmt,bigmt))
0de946608423 Uploaded
fubar
parents:
diff changeset
396 genesets = strsplit(genesets,'\t') # tabular. genesetid\tURLorwhatever\tgene_1\t..\tgene_n
0de946608423 Uploaded
fubar
parents:
diff changeset
397 outf = outfname
0de946608423 Uploaded
fubar
parents:
diff changeset
398 head=paste(myTitle,'edgeR GSEA')
0de946608423 Uploaded
fubar
parents:
diff changeset
399 write(head,file=outfname,append=F)
0de946608423 Uploaded
fubar
parents:
diff changeset
400 ntest=length(genesets)
0de946608423 Uploaded
fubar
parents:
diff changeset
401 urownames = toupper(rownames(y))
0de946608423 Uploaded
fubar
parents:
diff changeset
402 upcam = c()
0de946608423 Uploaded
fubar
parents:
diff changeset
403 downcam = c()
0de946608423 Uploaded
fubar
parents:
diff changeset
404 incam = c()
0de946608423 Uploaded
fubar
parents:
diff changeset
405 urls = c()
0de946608423 Uploaded
fubar
parents:
diff changeset
406 gsids = c()
0de946608423 Uploaded
fubar
parents:
diff changeset
407 for (i in 1:ntest) {
0de946608423 Uploaded
fubar
parents:
diff changeset
408 gs = unlist(genesets[i])
0de946608423 Uploaded
fubar
parents:
diff changeset
409 gsid = gs[1] # geneset_id
0de946608423 Uploaded
fubar
parents:
diff changeset
410 url = gs[2]
0de946608423 Uploaded
fubar
parents:
diff changeset
411 if (url > "") { url = paste("<a href=\'",url,"\'>",url,"</a>",sep="") }
0de946608423 Uploaded
fubar
parents:
diff changeset
412 glist = gs[3:length(gs)] # member gene symbols
0de946608423 Uploaded
fubar
parents:
diff changeset
413 glist = toupper(glist)
0de946608423 Uploaded
fubar
parents:
diff changeset
414 inglist = urownames %in% glist
0de946608423 Uploaded
fubar
parents:
diff changeset
415 nin = sum(inglist)
0de946608423 Uploaded
fubar
parents:
diff changeset
416 if ((nin > minnin) && (nin < maxnin)) {
0de946608423 Uploaded
fubar
parents:
diff changeset
417 incam = c(incam,inglist)
0de946608423 Uploaded
fubar
parents:
diff changeset
418 gsids = c(gsids,gsid)
0de946608423 Uploaded
fubar
parents:
diff changeset
419 urls = c(urls,url)
0de946608423 Uploaded
fubar
parents:
diff changeset
420 }
0de946608423 Uploaded
fubar
parents:
diff changeset
421 }
0de946608423 Uploaded
fubar
parents:
diff changeset
422 incam = as.list(incam)
0de946608423 Uploaded
fubar
parents:
diff changeset
423 names(incam) = gsids
0de946608423 Uploaded
fubar
parents:
diff changeset
424 allcam = camera(y=y,index=incam,design=design)
0de946608423 Uploaded
fubar
parents:
diff changeset
425 allcamres = cbind(geneset=gsids,allcam,URL=urls)
0de946608423 Uploaded
fubar
parents:
diff changeset
426 for (i in 1:ntest) {
0de946608423 Uploaded
fubar
parents:
diff changeset
427 camres = allcamres[i]
0de946608423 Uploaded
fubar
parents:
diff changeset
428 res = try(test = (camres\$Direction == "Up"))
0de946608423 Uploaded
fubar
parents:
diff changeset
429 if ("try-error" %in% class(res)) {
0de946608423 Uploaded
fubar
parents:
diff changeset
430 cat("test failed, camres = :")
0de946608423 Uploaded
fubar
parents:
diff changeset
431 print.noquote(camres)
0de946608423 Uploaded
fubar
parents:
diff changeset
432 } else { if (camres\$Direction == "Up")
0de946608423 Uploaded
fubar
parents:
diff changeset
433 { upcam = rbind(upcam,camres)
0de946608423 Uploaded
fubar
parents:
diff changeset
434 } else { downcam = rbind(downcam,camres)
0de946608423 Uploaded
fubar
parents:
diff changeset
435 }
0de946608423 Uploaded
fubar
parents:
diff changeset
436
0de946608423 Uploaded
fubar
parents:
diff changeset
437 }
0de946608423 Uploaded
fubar
parents:
diff changeset
438 }
0de946608423 Uploaded
fubar
parents:
diff changeset
439 uscam = upcam[order(upcam\$PValue),]
0de946608423 Uploaded
fubar
parents:
diff changeset
440 unadjp = uscam\$PValue
0de946608423 Uploaded
fubar
parents:
diff changeset
441 uscam\$adjPValue = p.adjust(unadjp,method=fdrtype)
0de946608423 Uploaded
fubar
parents:
diff changeset
442 nup = max(10,sum((uscam\$adjPValue < fdrthresh)))
0de946608423 Uploaded
fubar
parents:
diff changeset
443 dscam = downcam[order(downcam\$PValue),]
0de946608423 Uploaded
fubar
parents:
diff changeset
444 unadjp = dscam\$PValue
0de946608423 Uploaded
fubar
parents:
diff changeset
445 dscam\$adjPValue = p.adjust(unadjp,method=fdrtype)
0de946608423 Uploaded
fubar
parents:
diff changeset
446 ndown = max(10,sum((dscam\$adjPValue < fdrthresh)))
0de946608423 Uploaded
fubar
parents:
diff changeset
447 write.table(uscam,file=paste('camera_up',outfname,sep='_'),quote=F,sep='\t',row.names=F)
0de946608423 Uploaded
fubar
parents:
diff changeset
448 write.table(dscam,file=paste('camera_down',outfname,sep='_'),quote=F,sep='\t',row.names=F)
0de946608423 Uploaded
fubar
parents:
diff changeset
449 print.noquote(paste('@@@@@ Camera up top',nup,'gene sets:'))
0de946608423 Uploaded
fubar
parents:
diff changeset
450 write.table(head(uscam,nup),file="",quote=F,sep='\t',row.names=F)
0de946608423 Uploaded
fubar
parents:
diff changeset
451 print.noquote(paste('@@@@@ Camera down top',ndown,'gene sets:'))
0de946608423 Uploaded
fubar
parents:
diff changeset
452 write.table(head(dscam,ndown),file="",quote=F,sep='\t',row.names=F)
0de946608423 Uploaded
fubar
parents:
diff changeset
453 sink()
0de946608423 Uploaded
fubar
parents:
diff changeset
454 }
0de946608423 Uploaded
fubar
parents:
diff changeset
455
0de946608423 Uploaded
fubar
parents:
diff changeset
456
0de946608423 Uploaded
fubar
parents:
diff changeset
457 edgeIt = function (Count_Matrix=c(),group=c(),out_edgeR=F,out_Voom=F,out_DESeq2=F,fdrtype='fdr',priordf=5,
0de946608423 Uploaded
fubar
parents:
diff changeset
458 fdrthresh=0.05,outputdir='.', myTitle='Differential Counts',libSize=c(),useNDF=F,
0de946608423 Uploaded
fubar
parents:
diff changeset
459 filterquantile=0.2, subjects=c(),TreatmentName="Rx",ControlName="Ctrl",mydesign=NULL,
0de946608423 Uploaded
fubar
parents:
diff changeset
460 doDESeq2=T,doVoom=T,doCamera=T,doedgeR=T,org='hg19',
0de946608423 Uploaded
fubar
parents:
diff changeset
461 histgmt="", bigmt="/data/genomes/gsea/3.1/Abetterchoice_nocgp_c2_c3_c5_symbols_all.gmt",
0de946608423 Uploaded
fubar
parents:
diff changeset
462 doCook=F,DESeq_fitType="parameteric",robust_meth='ordinary')
0de946608423 Uploaded
fubar
parents:
diff changeset
463 {
0de946608423 Uploaded
fubar
parents:
diff changeset
464
0de946608423 Uploaded
fubar
parents:
diff changeset
465 logf = file('Differential.log', open = "a")
0de946608423 Uploaded
fubar
parents:
diff changeset
466 sink(logf,type = c("output", "message"))
0de946608423 Uploaded
fubar
parents:
diff changeset
467
0de946608423 Uploaded
fubar
parents:
diff changeset
468
0de946608423 Uploaded
fubar
parents:
diff changeset
469 run_edgeR = function(workCM,pdata,subjects,group,priordf,robust_meth,mydesign,mt,cmrowsums,out_edgeR,nonzerod)
0de946608423 Uploaded
fubar
parents:
diff changeset
470 {
0de946608423 Uploaded
fubar
parents:
diff changeset
471 logf = file('edgeR.log', open = "a")
0de946608423 Uploaded
fubar
parents:
diff changeset
472 sink(logf,type = c("output", "message"))
0de946608423 Uploaded
fubar
parents:
diff changeset
473 #### Setup myDGEList object
0de946608423 Uploaded
fubar
parents:
diff changeset
474 myDGEList = DGEList(counts=workCM, group = group)
0de946608423 Uploaded
fubar
parents:
diff changeset
475 myDGEList = calcNormFactors(myDGEList)
0de946608423 Uploaded
fubar
parents:
diff changeset
476 if (robust_meth == 'ordinary') {
0de946608423 Uploaded
fubar
parents:
diff changeset
477 myDGEList = estimateGLMCommonDisp(myDGEList,mydesign)
0de946608423 Uploaded
fubar
parents:
diff changeset
478 myDGEList = estimateGLMTrendedDisp(myDGEList,mydesign)
0de946608423 Uploaded
fubar
parents:
diff changeset
479 if (priordf > 0) { myDGEList = estimateGLMTagwiseDisp(myDGEList,mydesign,prior.df = priordf)
0de946608423 Uploaded
fubar
parents:
diff changeset
480 } else { myDGEList = estimateGLMTagwiseDisp(myDGEList,mydesign) }
0de946608423 Uploaded
fubar
parents:
diff changeset
481 comdisp = myDGEList\$common.dispersion
0de946608423 Uploaded
fubar
parents:
diff changeset
482 estpriorn = getPriorN(myDGEList)
0de946608423 Uploaded
fubar
parents:
diff changeset
483 print(paste("Common Dispersion =",comdisp,"CV = ",sqrt(comdisp),"getPriorN = ",estpriorn),quote=F)
0de946608423 Uploaded
fubar
parents:
diff changeset
484 } else {
0de946608423 Uploaded
fubar
parents:
diff changeset
485 myDGEList = estimateGLMRobustDisp(myDGEList,design=mydesign, prior.df = priordf, maxit = 6, residual.type = robust_meth)
0de946608423 Uploaded
fubar
parents:
diff changeset
486 }
0de946608423 Uploaded
fubar
parents:
diff changeset
487
0de946608423 Uploaded
fubar
parents:
diff changeset
488
0de946608423 Uploaded
fubar
parents:
diff changeset
489 DGLM = glmFit(myDGEList,design=mydesign)
0de946608423 Uploaded
fubar
parents:
diff changeset
490 DE = glmLRT(DGLM,coef=ncol(DGLM\$design)) # always last one - subject is first if needed
0de946608423 Uploaded
fubar
parents:
diff changeset
491 normData = cpm(myDGEList)
0de946608423 Uploaded
fubar
parents:
diff changeset
492 uoutput = cbind(
0de946608423 Uploaded
fubar
parents:
diff changeset
493 Name=as.character(rownames(myDGEList\$counts)),
0de946608423 Uploaded
fubar
parents:
diff changeset
494 DE\$table,
0de946608423 Uploaded
fubar
parents:
diff changeset
495 adj.p.value=p.adjust(DE\$table\$PValue, method=fdrtype),
0de946608423 Uploaded
fubar
parents:
diff changeset
496 Dispersion=myDGEList\$tagwise.dispersion,totreads=cmrowsums,normData,
0de946608423 Uploaded
fubar
parents:
diff changeset
497 myDGEList\$counts
0de946608423 Uploaded
fubar
parents:
diff changeset
498 )
0de946608423 Uploaded
fubar
parents:
diff changeset
499 soutput = uoutput[order(DE\$table\$PValue),] # sorted into p value order - for quick toptable
0de946608423 Uploaded
fubar
parents:
diff changeset
500 goodness = gof(DGLM, pcutoff=fdrthresh)
0de946608423 Uploaded
fubar
parents:
diff changeset
501 if (sum(goodness\$outlier) > 0) {
0de946608423 Uploaded
fubar
parents:
diff changeset
502 print.noquote('GLM outliers:')
0de946608423 Uploaded
fubar
parents:
diff changeset
503 print(paste(rownames(DGLM)[(goodness\$outlier)],collapse=','),quote=F)
0de946608423 Uploaded
fubar
parents:
diff changeset
504 } else {
0de946608423 Uploaded
fubar
parents:
diff changeset
505 print('No GLM fit outlier genes found\n')
0de946608423 Uploaded
fubar
parents:
diff changeset
506 }
0de946608423 Uploaded
fubar
parents:
diff changeset
507 z = limma::zscoreGamma(goodness\$gof.statistic, shape=goodness\$df/2, scale=2)
0de946608423 Uploaded
fubar
parents:
diff changeset
508 pdf(paste("edgeR",mt,"GoodnessofFit.pdf",sep='_'))
0de946608423 Uploaded
fubar
parents:
diff changeset
509 qq = qqnorm(z, panel.first=grid(), main="tagwise dispersion")
0de946608423 Uploaded
fubar
parents:
diff changeset
510 abline(0,1,lwd=3)
0de946608423 Uploaded
fubar
parents:
diff changeset
511 points(qq\$x[goodness\$outlier],qq\$y[goodness\$outlier], pch=16, col="maroon")
0de946608423 Uploaded
fubar
parents:
diff changeset
512 dev.off()
0de946608423 Uploaded
fubar
parents:
diff changeset
513 uniqueg = unique(group)
0de946608423 Uploaded
fubar
parents:
diff changeset
514 write.table(soutput,file=out_edgeR, quote=FALSE, sep="\t",row.names=F)
0de946608423 Uploaded
fubar
parents:
diff changeset
515 tt = cbind(
0de946608423 Uploaded
fubar
parents:
diff changeset
516 Name=as.character(rownames(myDGEList)),
0de946608423 Uploaded
fubar
parents:
diff changeset
517 DE\$table,
0de946608423 Uploaded
fubar
parents:
diff changeset
518 adj.p.value=p.adjust(DE\$table\$PValue, method=fdrtype),
0de946608423 Uploaded
fubar
parents:
diff changeset
519 Dispersion=myDGEList\$tagwise.dispersion,totreads=cmrowsums
0de946608423 Uploaded
fubar
parents:
diff changeset
520 )
0de946608423 Uploaded
fubar
parents:
diff changeset
521 tt = cbind(tt,URL=contigurls) # add to end so table isn't laid out strangely
0de946608423 Uploaded
fubar
parents:
diff changeset
522 stt = tt[order(DE\$table\$PValue),]
0de946608423 Uploaded
fubar
parents:
diff changeset
523 print.noquote("@@ edgeR Top tags\n")
0de946608423 Uploaded
fubar
parents:
diff changeset
524 print.noquote(stt[1:50,])
0de946608423 Uploaded
fubar
parents:
diff changeset
525 deTags = rownames(uoutput[uoutput\$adj.p.value < fdrthresh,])
0de946608423 Uploaded
fubar
parents:
diff changeset
526 nsig = length(deTags)
0de946608423 Uploaded
fubar
parents:
diff changeset
527 print.noquote(paste('@@',nsig,'tags significant at adj p=',fdrthresh))
0de946608423 Uploaded
fubar
parents:
diff changeset
528 deColours = ifelse(deTags,'red','black')
0de946608423 Uploaded
fubar
parents:
diff changeset
529 pdf(paste("edgeR",mt,"BCV_vs_abundance.pdf",sep="_"))
0de946608423 Uploaded
fubar
parents:
diff changeset
530 plotBCV(myDGEList, cex=0.3, main="Biological CV vs abundance")
0de946608423 Uploaded
fubar
parents:
diff changeset
531 dev.off()
0de946608423 Uploaded
fubar
parents:
diff changeset
532 dg = myDGEList[order(DE\$table\$PValue),]
0de946608423 Uploaded
fubar
parents:
diff changeset
533 outpdfname= paste("edgeR",mt,"top_100_heatmap.pdf",sep="_")
0de946608423 Uploaded
fubar
parents:
diff changeset
534 ocpm = normData[order(DE\$table\$PValue),]
0de946608423 Uploaded
fubar
parents:
diff changeset
535 ocpm = ocpm[c(1:100),]
0de946608423 Uploaded
fubar
parents:
diff changeset
536 hmap2(ocpm,TName=TName,group=group,outpdfname=outpdfname,myTitle=paste(myTitle,'Heatmap'))
0de946608423 Uploaded
fubar
parents:
diff changeset
537 outSmear = paste("edgeR",mt,"smearplot.pdf",sep="_")
0de946608423 Uploaded
fubar
parents:
diff changeset
538 outMain = paste("Smear Plot for ",TName,' Vs ',CName,' (FDR@',fdrthresh,' N = ',nsig,')',sep='')
0de946608423 Uploaded
fubar
parents:
diff changeset
539 smearPlot(myDGEList=myDGEList,deTags=deTags, outSmear=outSmear, outMain = outMain)
0de946608423 Uploaded
fubar
parents:
diff changeset
540 qqPlot(descr=paste(myTitle,'edgeR adj p QQ plot'),pvector=tt\$adj.p.value,outpdf=paste('edgeR',mt,'qqplot.pdf',sep='_'))
0de946608423 Uploaded
fubar
parents:
diff changeset
541 topresults.edgeR = soutput[which(soutput\$adj.p.value < fdrthresh), ]
0de946608423 Uploaded
fubar
parents:
diff changeset
542 edgeRcountsindex = which(allgenes %in% rownames(topresults.edgeR))
0de946608423 Uploaded
fubar
parents:
diff changeset
543 edgeRcounts = rep(0, length(allgenes))
0de946608423 Uploaded
fubar
parents:
diff changeset
544 edgeRcounts[edgeRcountsindex] = 1 # Create venn diagram of hits
0de946608423 Uploaded
fubar
parents:
diff changeset
545 sink()
0de946608423 Uploaded
fubar
parents:
diff changeset
546 return(list(myDGEList=myDGEList,edgeRcounts=edgeRcounts))
0de946608423 Uploaded
fubar
parents:
diff changeset
547 } ### run_edgeR
0de946608423 Uploaded
fubar
parents:
diff changeset
548
0de946608423 Uploaded
fubar
parents:
diff changeset
549
0de946608423 Uploaded
fubar
parents:
diff changeset
550 run_DESeq2 = function(workCM,pdata,subjects,group,out_DESeq2,mt,DESeq_fitType)
0de946608423 Uploaded
fubar
parents:
diff changeset
551
0de946608423 Uploaded
fubar
parents:
diff changeset
552 {
0de946608423 Uploaded
fubar
parents:
diff changeset
553 logf = file("DESeq2.log", open = "a")
0de946608423 Uploaded
fubar
parents:
diff changeset
554 sink(logf,type = c("output", "message"))
0de946608423 Uploaded
fubar
parents:
diff changeset
555 # DESeq2
0de946608423 Uploaded
fubar
parents:
diff changeset
556 require('DESeq2')
0de946608423 Uploaded
fubar
parents:
diff changeset
557 library('RColorBrewer')
0de946608423 Uploaded
fubar
parents:
diff changeset
558 if (length(subjects) == 0)
0de946608423 Uploaded
fubar
parents:
diff changeset
559 {
0de946608423 Uploaded
fubar
parents:
diff changeset
560 pdata = data.frame(Name=colnames(workCM),Rx=group,row.names=colnames(workCM))
0de946608423 Uploaded
fubar
parents:
diff changeset
561 deSEQds = DESeqDataSetFromMatrix(countData = workCM, colData = pdata, design = formula(~ Rx))
0de946608423 Uploaded
fubar
parents:
diff changeset
562 } else {
0de946608423 Uploaded
fubar
parents:
diff changeset
563 pdata = data.frame(Name=colnames(workCM),Rx=group,subjects=subjects,row.names=colnames(workCM))
0de946608423 Uploaded
fubar
parents:
diff changeset
564 deSEQds = DESeqDataSetFromMatrix(countData = workCM, colData = pdata, design = formula(~ subjects + Rx))
0de946608423 Uploaded
fubar
parents:
diff changeset
565 }
0de946608423 Uploaded
fubar
parents:
diff changeset
566 deSeqDatsizefac = estimateSizeFactors(deSEQds)
0de946608423 Uploaded
fubar
parents:
diff changeset
567 deSeqDatdisp = estimateDispersions(deSeqDatsizefac,fitType=DESeq_fitType)
0de946608423 Uploaded
fubar
parents:
diff changeset
568 resDESeq = nbinomWaldTest(deSeqDatdisp)
0de946608423 Uploaded
fubar
parents:
diff changeset
569 rDESeq = as.data.frame(results(resDESeq))
0de946608423 Uploaded
fubar
parents:
diff changeset
570 rDESeq = cbind(Contig=rownames(workCM),rDESeq,NReads=cmrowsums,URL=contigurls)
0de946608423 Uploaded
fubar
parents:
diff changeset
571 srDESeq = rDESeq[order(rDESeq\$pvalue),]
0de946608423 Uploaded
fubar
parents:
diff changeset
572 qqPlot(descr=paste(myTitle,'DESeq2 adj p qq plot'),pvector=rDESeq\$padj,outpdf=paste('DESeq2',mt,'qqplot.pdf',sep="_"))
0de946608423 Uploaded
fubar
parents:
diff changeset
573 cat("# DESeq top 50\n")
0de946608423 Uploaded
fubar
parents:
diff changeset
574 print.noquote(srDESeq[1:50,])
0de946608423 Uploaded
fubar
parents:
diff changeset
575 write.table(srDESeq,file=out_DESeq2, quote=FALSE, sep="\t",row.names=F)
0de946608423 Uploaded
fubar
parents:
diff changeset
576 topresults.DESeq = rDESeq[which(rDESeq\$padj < fdrthresh), ]
0de946608423 Uploaded
fubar
parents:
diff changeset
577 DESeqcountsindex = which(allgenes %in% rownames(topresults.DESeq))
0de946608423 Uploaded
fubar
parents:
diff changeset
578 DESeqcounts = rep(0, length(allgenes))
0de946608423 Uploaded
fubar
parents:
diff changeset
579 DESeqcounts[DESeqcountsindex] = 1
0de946608423 Uploaded
fubar
parents:
diff changeset
580 pdf(paste("DESeq2",mt,"dispersion_estimates.pdf",sep='_'))
0de946608423 Uploaded
fubar
parents:
diff changeset
581 plotDispEsts(resDESeq)
0de946608423 Uploaded
fubar
parents:
diff changeset
582 dev.off()
0de946608423 Uploaded
fubar
parents:
diff changeset
583 ysmall = abs(min(rDESeq\$log2FoldChange))
0de946608423 Uploaded
fubar
parents:
diff changeset
584 ybig = abs(max(rDESeq\$log2FoldChange))
0de946608423 Uploaded
fubar
parents:
diff changeset
585 ylimit = min(4,ysmall,ybig)
0de946608423 Uploaded
fubar
parents:
diff changeset
586 pdf(paste("DESeq2",mt,"MA_plot.pdf",sep="_"))
0de946608423 Uploaded
fubar
parents:
diff changeset
587 plotMA(resDESeq,main=paste(myTitle,"DESeq2 MA plot"),ylim=c(-ylimit,ylimit))
0de946608423 Uploaded
fubar
parents:
diff changeset
588 dev.off()
0de946608423 Uploaded
fubar
parents:
diff changeset
589 rlogres = rlogTransformation(resDESeq)
0de946608423 Uploaded
fubar
parents:
diff changeset
590 sampledists = dist( t( assay(rlogres) ) )
0de946608423 Uploaded
fubar
parents:
diff changeset
591 sdmat = as.matrix(sampledists)
0de946608423 Uploaded
fubar
parents:
diff changeset
592 pdf(paste("DESeq2",mt,"sample_distance_plot.pdf",sep="_"))
0de946608423 Uploaded
fubar
parents:
diff changeset
593 heatmap.2(sdmat,trace="none",main=paste(myTitle,"DESeq2 sample distances"),
0de946608423 Uploaded
fubar
parents:
diff changeset
594 col = colorRampPalette( rev(brewer.pal(9, "RdBu")) )(255))
0de946608423 Uploaded
fubar
parents:
diff changeset
595 dev.off()
0de946608423 Uploaded
fubar
parents:
diff changeset
596 result = try( (ppca = plotPCA( varianceStabilizingTransformation(deSeqDatdisp,blind=T), intgroup=c("Rx","Name")) ) )
0de946608423 Uploaded
fubar
parents:
diff changeset
597 if ("try-error" %in% class(result)) {
0de946608423 Uploaded
fubar
parents:
diff changeset
598 print.noquote('DESeq2 plotPCA failed.')
0de946608423 Uploaded
fubar
parents:
diff changeset
599 } else {
0de946608423 Uploaded
fubar
parents:
diff changeset
600 pdf(paste("DESeq2",mt,"PCA_plot.pdf",sep="_"))
0de946608423 Uploaded
fubar
parents:
diff changeset
601 #### wtf - print? Seems needed to get this to work
0de946608423 Uploaded
fubar
parents:
diff changeset
602 print(ppca)
0de946608423 Uploaded
fubar
parents:
diff changeset
603 dev.off()
0de946608423 Uploaded
fubar
parents:
diff changeset
604 }
0de946608423 Uploaded
fubar
parents:
diff changeset
605 sink()
0de946608423 Uploaded
fubar
parents:
diff changeset
606 return(DESeqcounts)
0de946608423 Uploaded
fubar
parents:
diff changeset
607 }
0de946608423 Uploaded
fubar
parents:
diff changeset
608
0de946608423 Uploaded
fubar
parents:
diff changeset
609
0de946608423 Uploaded
fubar
parents:
diff changeset
610 run_Voom = function(workCM,pdata,subjects,group,mydesign,mt,out_Voom)
0de946608423 Uploaded
fubar
parents:
diff changeset
611 {
0de946608423 Uploaded
fubar
parents:
diff changeset
612 logf = file('VOOM.log', open = "a")
0de946608423 Uploaded
fubar
parents:
diff changeset
613 sink(logf,type = c("output", "message"))
0de946608423 Uploaded
fubar
parents:
diff changeset
614 if (doedgeR == F) {
0de946608423 Uploaded
fubar
parents:
diff changeset
615 #### Setup myDGEList object
0de946608423 Uploaded
fubar
parents:
diff changeset
616 myDGEList = DGEList(counts=workCM, group = group)
0de946608423 Uploaded
fubar
parents:
diff changeset
617 myDGEList = calcNormFactors(myDGEList)
0de946608423 Uploaded
fubar
parents:
diff changeset
618 myDGEList = estimateGLMCommonDisp(myDGEList,mydesign)
0de946608423 Uploaded
fubar
parents:
diff changeset
619 myDGEList = estimateGLMTrendedDisp(myDGEList,mydesign)
0de946608423 Uploaded
fubar
parents:
diff changeset
620 myDGEList = estimateGLMTagwiseDisp(myDGEList,mydesign)
0de946608423 Uploaded
fubar
parents:
diff changeset
621 }
0de946608423 Uploaded
fubar
parents:
diff changeset
622 pdf(paste("VOOM",mt,"mean_variance_plot.pdf",sep='_'))
0de946608423 Uploaded
fubar
parents:
diff changeset
623 dat.voomed <- voom(myDGEList, mydesign, plot = TRUE, normalize.method="quantil", lib.size = NULL)
0de946608423 Uploaded
fubar
parents:
diff changeset
624 dev.off()
0de946608423 Uploaded
fubar
parents:
diff changeset
625 # Use limma to fit data
0de946608423 Uploaded
fubar
parents:
diff changeset
626 fit = lmFit(dat.voomed, mydesign)
0de946608423 Uploaded
fubar
parents:
diff changeset
627 fit = eBayes(fit)
0de946608423 Uploaded
fubar
parents:
diff changeset
628 rvoom = topTable(fit, coef = length(colnames(mydesign)), adj = fdrtype, n = Inf, sort="none")
0de946608423 Uploaded
fubar
parents:
diff changeset
629 qqPlot(descr=paste(myTitle,'VOOM-limma adj p QQ plot'),pvector=rvoom\$adj.P.Val,outpdf=paste('VOOM',mt,'qqplot.pdf',sep='_'))
0de946608423 Uploaded
fubar
parents:
diff changeset
630 rownames(rvoom) = rownames(workCM)
0de946608423 Uploaded
fubar
parents:
diff changeset
631 rvoom = cbind(Contig=rownames(workCM),rvoom,NReads=cmrowsums,URL=contigurls)
0de946608423 Uploaded
fubar
parents:
diff changeset
632 srvoom = rvoom[order(rvoom\$P.Value),]
0de946608423 Uploaded
fubar
parents:
diff changeset
633 cat("# VOOM top 50\n")
0de946608423 Uploaded
fubar
parents:
diff changeset
634 print(srvoom[1:50,])
0de946608423 Uploaded
fubar
parents:
diff changeset
635 write.table(srvoom,file=out_Voom, quote=FALSE, sep="\t",row.names=F)
0de946608423 Uploaded
fubar
parents:
diff changeset
636 # Use an FDR cutoff to find interesting samples for edgeR, DESeq and voom/limma
0de946608423 Uploaded
fubar
parents:
diff changeset
637 topresults.voom = rvoom[which(rvoom\$adj.P.Val < fdrthresh), ]
0de946608423 Uploaded
fubar
parents:
diff changeset
638 voomcountsindex <- which(allgenes %in% rownames(topresults.voom))
0de946608423 Uploaded
fubar
parents:
diff changeset
639 voomcounts = rep(0, length(allgenes))
0de946608423 Uploaded
fubar
parents:
diff changeset
640 voomcounts[voomcountsindex] = 1
0de946608423 Uploaded
fubar
parents:
diff changeset
641 sink()
0de946608423 Uploaded
fubar
parents:
diff changeset
642 return(voomcounts)
0de946608423 Uploaded
fubar
parents:
diff changeset
643 }
0de946608423 Uploaded
fubar
parents:
diff changeset
644
0de946608423 Uploaded
fubar
parents:
diff changeset
645
0de946608423 Uploaded
fubar
parents:
diff changeset
646 #### data cleaning and analsis control starts here
0de946608423 Uploaded
fubar
parents:
diff changeset
647
0de946608423 Uploaded
fubar
parents:
diff changeset
648
0de946608423 Uploaded
fubar
parents:
diff changeset
649 # Error handling
0de946608423 Uploaded
fubar
parents:
diff changeset
650 nugroup = length(unique(group))
0de946608423 Uploaded
fubar
parents:
diff changeset
651 if (nugroup!=2){
0de946608423 Uploaded
fubar
parents:
diff changeset
652 print("Number of conditions identified in experiment does not equal 2")
0de946608423 Uploaded
fubar
parents:
diff changeset
653 q()
0de946608423 Uploaded
fubar
parents:
diff changeset
654 }
0de946608423 Uploaded
fubar
parents:
diff changeset
655 require(edgeR)
0de946608423 Uploaded
fubar
parents:
diff changeset
656 options(width = 512)
0de946608423 Uploaded
fubar
parents:
diff changeset
657 mt = paste(unlist(strsplit(myTitle,'_')),collapse=" ")
0de946608423 Uploaded
fubar
parents:
diff changeset
658 allN = nrow(Count_Matrix)
0de946608423 Uploaded
fubar
parents:
diff changeset
659 nscut = round(ncol(Count_Matrix)/2) # half samples
0de946608423 Uploaded
fubar
parents:
diff changeset
660 colTotmillionreads = colSums(Count_Matrix)/1e6
0de946608423 Uploaded
fubar
parents:
diff changeset
661 counts.dataframe = as.data.frame(c())
0de946608423 Uploaded
fubar
parents:
diff changeset
662 rawrs = rowSums(Count_Matrix)
0de946608423 Uploaded
fubar
parents:
diff changeset
663 nonzerod = Count_Matrix[(rawrs > 0),] # remove all zero count genes
0de946608423 Uploaded
fubar
parents:
diff changeset
664 nzN = nrow(nonzerod)
0de946608423 Uploaded
fubar
parents:
diff changeset
665 nzrs = rowSums(nonzerod)
0de946608423 Uploaded
fubar
parents:
diff changeset
666 zN = allN - nzN
0de946608423 Uploaded
fubar
parents:
diff changeset
667 print('@@@ Quantiles for non-zero row counts:',quote=F)
0de946608423 Uploaded
fubar
parents:
diff changeset
668 print(quantile(nzrs,probs=seq(0,1,0.1)),quote=F)
0de946608423 Uploaded
fubar
parents:
diff changeset
669 if (useNDF == T)
0de946608423 Uploaded
fubar
parents:
diff changeset
670 {
0de946608423 Uploaded
fubar
parents:
diff changeset
671 gt1rpin3 = rowSums(Count_Matrix/expandAsMatrix(colTotmillionreads,dim(Count_Matrix)) >= 1) >= nscut
0de946608423 Uploaded
fubar
parents:
diff changeset
672 lo = colSums(Count_Matrix[!gt1rpin3,])
0de946608423 Uploaded
fubar
parents:
diff changeset
673 workCM = Count_Matrix[gt1rpin3,]
0de946608423 Uploaded
fubar
parents:
diff changeset
674 cleanrs = rowSums(workCM)
0de946608423 Uploaded
fubar
parents:
diff changeset
675 cleanN = length(cleanrs)
0de946608423 Uploaded
fubar
parents:
diff changeset
676 meth = paste( "After removing",length(lo),"contigs with fewer than ",nscut," sample read counts >= 1 per million, there are",sep="")
0de946608423 Uploaded
fubar
parents:
diff changeset
677 print(paste("Read",allN,"contigs. Removed",zN,"contigs with no reads.",meth,cleanN,"contigs"),quote=F)
0de946608423 Uploaded
fubar
parents:
diff changeset
678 maint = paste('Filter >=1/million reads in >=',nscut,'samples')
0de946608423 Uploaded
fubar
parents:
diff changeset
679 } else {
0de946608423 Uploaded
fubar
parents:
diff changeset
680 useme = (nzrs > quantile(nzrs,filterquantile))
0de946608423 Uploaded
fubar
parents:
diff changeset
681 workCM = nonzerod[useme,]
0de946608423 Uploaded
fubar
parents:
diff changeset
682 lo = colSums(nonzerod[!useme,])
0de946608423 Uploaded
fubar
parents:
diff changeset
683 cleanrs = rowSums(workCM)
0de946608423 Uploaded
fubar
parents:
diff changeset
684 cleanN = length(cleanrs)
0de946608423 Uploaded
fubar
parents:
diff changeset
685 meth = paste("After filtering at count quantile =",filterquantile,", there are",sep="")
0de946608423 Uploaded
fubar
parents:
diff changeset
686 print(paste('Read',allN,"contigs. Removed",zN,"with no reads.",meth,cleanN,"contigs"),quote=F)
0de946608423 Uploaded
fubar
parents:
diff changeset
687 maint = paste('Filter below',filterquantile,'quantile')
0de946608423 Uploaded
fubar
parents:
diff changeset
688 }
0de946608423 Uploaded
fubar
parents:
diff changeset
689 cumPlot(rawrs=rawrs,cleanrs=cleanrs,maint=maint,myTitle=myTitle)
0de946608423 Uploaded
fubar
parents:
diff changeset
690 allgenes = rownames(workCM)
0de946608423 Uploaded
fubar
parents:
diff changeset
691 reg = "^chr([0-9]+):([0-9]+)-([0-9]+)" # ucsc chr:start-end regexp
0de946608423 Uploaded
fubar
parents:
diff changeset
692 genecards="<a href=\'http://www.genecards.org/index.php?path=/Search/keyword/"
0de946608423 Uploaded
fubar
parents:
diff changeset
693 ucsc = paste("<a href=\'http://genome.ucsc.edu/cgi-bin/hgTracks?db=",org,sep='')
0de946608423 Uploaded
fubar
parents:
diff changeset
694 testreg = str_match(allgenes,reg)
0de946608423 Uploaded
fubar
parents:
diff changeset
695 if (sum(!is.na(testreg[,1]))/length(testreg[,1]) > 0.8) # is ucsc style string
0de946608423 Uploaded
fubar
parents:
diff changeset
696 {
0de946608423 Uploaded
fubar
parents:
diff changeset
697 print("@@ using ucsc substitution for urls")
0de946608423 Uploaded
fubar
parents:
diff changeset
698 contigurls = paste0(ucsc,"&amp;position=chr",testreg[,2],":",testreg[,3],"-",testreg[,4],"\'>",allgenes,"</a>")
0de946608423 Uploaded
fubar
parents:
diff changeset
699 } else {
0de946608423 Uploaded
fubar
parents:
diff changeset
700 print("@@ using genecards substitution for urls")
0de946608423 Uploaded
fubar
parents:
diff changeset
701 contigurls = paste0(genecards,allgenes,"\'>",allgenes,"</a>")
0de946608423 Uploaded
fubar
parents:
diff changeset
702 }
0de946608423 Uploaded
fubar
parents:
diff changeset
703 print.noquote(paste("@@ Total low count contigs per sample = ",paste(table(lo),collapse=',')))
0de946608423 Uploaded
fubar
parents:
diff changeset
704 cmrowsums = rowSums(workCM)
0de946608423 Uploaded
fubar
parents:
diff changeset
705 TName=unique(group)[1]
0de946608423 Uploaded
fubar
parents:
diff changeset
706 CName=unique(group)[2]
0de946608423 Uploaded
fubar
parents:
diff changeset
707 if (is.null(mydesign)) {
0de946608423 Uploaded
fubar
parents:
diff changeset
708 if (length(subjects) == 0)
0de946608423 Uploaded
fubar
parents:
diff changeset
709 {
0de946608423 Uploaded
fubar
parents:
diff changeset
710 mydesign = model.matrix(~group)
0de946608423 Uploaded
fubar
parents:
diff changeset
711 }
0de946608423 Uploaded
fubar
parents:
diff changeset
712 else {
0de946608423 Uploaded
fubar
parents:
diff changeset
713 subjf = factor(subjects)
0de946608423 Uploaded
fubar
parents:
diff changeset
714 mydesign = model.matrix(~subjf+group) # we block on subject so make group last to simplify finding it
0de946608423 Uploaded
fubar
parents:
diff changeset
715 }
0de946608423 Uploaded
fubar
parents:
diff changeset
716 }
0de946608423 Uploaded
fubar
parents:
diff changeset
717 print.noquote(paste('Using samples:',paste(colnames(workCM),collapse=',')))
0de946608423 Uploaded
fubar
parents:
diff changeset
718 print.noquote('Using design matrix:')
0de946608423 Uploaded
fubar
parents:
diff changeset
719 print.noquote(mydesign)
0de946608423 Uploaded
fubar
parents:
diff changeset
720 normData = cpm(workCM)*1e6
0de946608423 Uploaded
fubar
parents:
diff changeset
721 colnames(normData) = paste( colnames(workCM),'N',sep="_")
0de946608423 Uploaded
fubar
parents:
diff changeset
722 print(paste('Raw sample read totals',paste(colSums(nonzerod,na.rm=T),collapse=',')))
0de946608423 Uploaded
fubar
parents:
diff changeset
723
0de946608423 Uploaded
fubar
parents:
diff changeset
724 if (doedgeR == T) {
0de946608423 Uploaded
fubar
parents:
diff changeset
725 eres = run_edgeR(workCM,pdata,subjects,group,priordf,robust_meth,mydesign,mt,cmrowsums,out_edgeR,nonzerod)
0de946608423 Uploaded
fubar
parents:
diff changeset
726 myDGEList = eres\$myDGEList
0de946608423 Uploaded
fubar
parents:
diff changeset
727 edgeRcounts = eres\$edgeRcounts
0de946608423 Uploaded
fubar
parents:
diff changeset
728 #### Plot MDS
0de946608423 Uploaded
fubar
parents:
diff changeset
729 sample_colors = match(group,levels(group))
0de946608423 Uploaded
fubar
parents:
diff changeset
730 sampleTypes = levels(factor(group))
0de946608423 Uploaded
fubar
parents:
diff changeset
731 print.noquote(sampleTypes)
0de946608423 Uploaded
fubar
parents:
diff changeset
732 pdf(paste("edgeR",mt,"MDSplot.pdf",sep='_'))
0de946608423 Uploaded
fubar
parents:
diff changeset
733 plotMDS.DGEList(myDGEList,main=paste("MDS for",myTitle),cex=0.5,col=sample_colors,pch=sample_colors)
0de946608423 Uploaded
fubar
parents:
diff changeset
734 legend(x="topleft", legend = sampleTypes,col=c(1:length(sampleTypes)), pch=19)
0de946608423 Uploaded
fubar
parents:
diff changeset
735 grid(col="blue")
0de946608423 Uploaded
fubar
parents:
diff changeset
736 dev.off()
0de946608423 Uploaded
fubar
parents:
diff changeset
737 scale <- myDGEList\$samples\$lib.size*myDGEList\$samples\$norm.factors
0de946608423 Uploaded
fubar
parents:
diff changeset
738 normCounts <- round(t(t(myDGEList\$counts)/scale)*mean(scale))
0de946608423 Uploaded
fubar
parents:
diff changeset
739 try({boxPlot(rawrs=nzd,cleanrs=log2(normCounts+1),maint='Effects of TMM size normalisation',myTitle=myTitle,pdfname=paste("edgeR",mt,"raw_norm_counts_box.pdf",sep='_'))},T)
0de946608423 Uploaded
fubar
parents:
diff changeset
740 }
0de946608423 Uploaded
fubar
parents:
diff changeset
741 if (doDESeq2 == T) { DESeqcounts = run_DESeq2(workCM,pdata,subjects,group,out_DESeq2,mt,DESeq_fitType) }
0de946608423 Uploaded
fubar
parents:
diff changeset
742 if (doVoom == T) { voomcounts = run_Voom(workCM,pdata,subjects,group,mydesign,mt,out_Voom) }
0de946608423 Uploaded
fubar
parents:
diff changeset
743
0de946608423 Uploaded
fubar
parents:
diff changeset
744
0de946608423 Uploaded
fubar
parents:
diff changeset
745 if (doCamera) {
0de946608423 Uploaded
fubar
parents:
diff changeset
746 doGSEA(y=myDGEList,design=mydesign,histgmt=histgmt,bigmt=bigmt,ntest=20,myTitle=myTitle,
0de946608423 Uploaded
fubar
parents:
diff changeset
747 outfname=paste("GSEA_Camera",mt,"table.xls",sep="_"),fdrthresh=fdrthresh,fdrtype=fdrtype)
0de946608423 Uploaded
fubar
parents:
diff changeset
748 }
0de946608423 Uploaded
fubar
parents:
diff changeset
749 counts.dataframe = c()
0de946608423 Uploaded
fubar
parents:
diff changeset
750 vennmain = 'no venn'
0de946608423 Uploaded
fubar
parents:
diff changeset
751 if ((doDESeq2==T) || (doVoom==T) || (doedgeR==T)) {
0de946608423 Uploaded
fubar
parents:
diff changeset
752 if ((doVoom==T) && (doDESeq2==T) && (doedgeR==T)) {
0de946608423 Uploaded
fubar
parents:
diff changeset
753 vennmain = paste(mt,'Voom,edgeR and DESeq2 overlap at FDR=',fdrthresh)
0de946608423 Uploaded
fubar
parents:
diff changeset
754 counts.dataframe = data.frame(edgeR = edgeRcounts, DESeq2 = DESeqcounts,
0de946608423 Uploaded
fubar
parents:
diff changeset
755 VOOM_limma = voomcounts, row.names = allgenes)
0de946608423 Uploaded
fubar
parents:
diff changeset
756 } else if ((doDESeq2==T) && (doedgeR==T)) {
0de946608423 Uploaded
fubar
parents:
diff changeset
757 vennmain = paste(mt,'DESeq2 and edgeR overlap at FDR=',fdrthresh)
0de946608423 Uploaded
fubar
parents:
diff changeset
758 counts.dataframe = data.frame(edgeR = edgeRcounts, DESeq2 = DESeqcounts, row.names = allgenes)
0de946608423 Uploaded
fubar
parents:
diff changeset
759 } else if ((doVoom==T) && (doedgeR==T)) {
0de946608423 Uploaded
fubar
parents:
diff changeset
760 vennmain = paste(mt,'Voom and edgeR overlap at FDR=',fdrthresh)
0de946608423 Uploaded
fubar
parents:
diff changeset
761 counts.dataframe = data.frame(edgeR = edgeRcounts, VOOM_limma = voomcounts, row.names = allgenes)
0de946608423 Uploaded
fubar
parents:
diff changeset
762 }
0de946608423 Uploaded
fubar
parents:
diff changeset
763
0de946608423 Uploaded
fubar
parents:
diff changeset
764 if (nrow(counts.dataframe > 1)) {
0de946608423 Uploaded
fubar
parents:
diff changeset
765 counts.venn = vennCounts(counts.dataframe)
0de946608423 Uploaded
fubar
parents:
diff changeset
766 vennf = paste("Differential_venn",mt,"significant_genes_overlap.pdf",sep="_")
0de946608423 Uploaded
fubar
parents:
diff changeset
767 pdf(vennf)
0de946608423 Uploaded
fubar
parents:
diff changeset
768 vennDiagram(counts.venn,main=vennmain,col="maroon")
0de946608423 Uploaded
fubar
parents:
diff changeset
769 dev.off()
0de946608423 Uploaded
fubar
parents:
diff changeset
770 }
0de946608423 Uploaded
fubar
parents:
diff changeset
771 } #### doDESeq2 or doVoom
0de946608423 Uploaded
fubar
parents:
diff changeset
772 sink()
0de946608423 Uploaded
fubar
parents:
diff changeset
773 }
0de946608423 Uploaded
fubar
parents:
diff changeset
774 #### Done
0de946608423 Uploaded
fubar
parents:
diff changeset
775 ]]>
0de946608423 Uploaded
fubar
parents:
diff changeset
776 builtin_gmt = ""
0de946608423 Uploaded
fubar
parents:
diff changeset
777 history_gmt = ""
0de946608423 Uploaded
fubar
parents:
diff changeset
778 history_gmt_name = ""
0de946608423 Uploaded
fubar
parents:
diff changeset
779 out_edgeR = F
0de946608423 Uploaded
fubar
parents:
diff changeset
780 out_DESeq2 = F
0de946608423 Uploaded
fubar
parents:
diff changeset
781 out_Voom = "$out_VOOM"
0de946608423 Uploaded
fubar
parents:
diff changeset
782 edgeR_robust_meth = "ordinary"
0de946608423 Uploaded
fubar
parents:
diff changeset
783 doDESeq2 = $DESeq2.doDESeq2
0de946608423 Uploaded
fubar
parents:
diff changeset
784 doVoom = $doVoom
0de946608423 Uploaded
fubar
parents:
diff changeset
785 doCamera = F
0de946608423 Uploaded
fubar
parents:
diff changeset
786 doedgeR = $edgeR.doedgeR
0de946608423 Uploaded
fubar
parents:
diff changeset
787 edgeR_priordf = 10
0de946608423 Uploaded
fubar
parents:
diff changeset
788
0de946608423 Uploaded
fubar
parents:
diff changeset
789
0de946608423 Uploaded
fubar
parents:
diff changeset
790 #if $doVoom == "T":
0de946608423 Uploaded
fubar
parents:
diff changeset
791 out_Voom = "$out_VOOM"
0de946608423 Uploaded
fubar
parents:
diff changeset
792 #end if
0de946608423 Uploaded
fubar
parents:
diff changeset
793
0de946608423 Uploaded
fubar
parents:
diff changeset
794 #if $DESeq2.doDESeq2 == "T":
0de946608423 Uploaded
fubar
parents:
diff changeset
795 out_DESeq2 = "$out_DESeq2"
0de946608423 Uploaded
fubar
parents:
diff changeset
796 doDESeq2 = T
0de946608423 Uploaded
fubar
parents:
diff changeset
797 DESeq_fitType = "$DESeq2.DESeq_fitType"
0de946608423 Uploaded
fubar
parents:
diff changeset
798 #end if
0de946608423 Uploaded
fubar
parents:
diff changeset
799
0de946608423 Uploaded
fubar
parents:
diff changeset
800 #if $edgeR.doedgeR == "T":
0de946608423 Uploaded
fubar
parents:
diff changeset
801 out_edgeR = "$out_edgeR"
0de946608423 Uploaded
fubar
parents:
diff changeset
802 edgeR_priordf = $edgeR.edgeR_priordf
0de946608423 Uploaded
fubar
parents:
diff changeset
803 edgeR_robust_meth = "$edgeR.edgeR_robust_method"
0de946608423 Uploaded
fubar
parents:
diff changeset
804 #end if
0de946608423 Uploaded
fubar
parents:
diff changeset
805
0de946608423 Uploaded
fubar
parents:
diff changeset
806
0de946608423 Uploaded
fubar
parents:
diff changeset
807 if (sum(c(doedgeR,doVoom,doDESeq2)) == 0)
0de946608423 Uploaded
fubar
parents:
diff changeset
808 {
0de946608423 Uploaded
fubar
parents:
diff changeset
809 write("No methods chosen - nothing to do! Please try again after choosing one or more methods", stderr())
0de946608423 Uploaded
fubar
parents:
diff changeset
810 quit(save="no",status=2)
0de946608423 Uploaded
fubar
parents:
diff changeset
811 }
0de946608423 Uploaded
fubar
parents:
diff changeset
812
0de946608423 Uploaded
fubar
parents:
diff changeset
813 Out_Dir = "$html_file.files_path"
0de946608423 Uploaded
fubar
parents:
diff changeset
814 Input = "$input1"
0de946608423 Uploaded
fubar
parents:
diff changeset
815 TreatmentName = "$treatment_name"
0de946608423 Uploaded
fubar
parents:
diff changeset
816 TreatmentCols = "$Treat_cols"
0de946608423 Uploaded
fubar
parents:
diff changeset
817 ControlName = "$control_name"
0de946608423 Uploaded
fubar
parents:
diff changeset
818 ControlCols= "$Control_cols"
0de946608423 Uploaded
fubar
parents:
diff changeset
819 org = "$input1.dbkey"
0de946608423 Uploaded
fubar
parents:
diff changeset
820 if (org == "") { org = "hg19"}
0de946608423 Uploaded
fubar
parents:
diff changeset
821 fdrtype = "$fdrtype"
0de946608423 Uploaded
fubar
parents:
diff changeset
822 fdrthresh = $fdrthresh
0de946608423 Uploaded
fubar
parents:
diff changeset
823 useNDF = $useNDF
0de946608423 Uploaded
fubar
parents:
diff changeset
824 fQ = $fQ # non-differential centile cutoff
0de946608423 Uploaded
fubar
parents:
diff changeset
825 myTitle = "$title"
0de946608423 Uploaded
fubar
parents:
diff changeset
826 sids = strsplit("$subjectids",',')
0de946608423 Uploaded
fubar
parents:
diff changeset
827 subjects = unlist(sids)
0de946608423 Uploaded
fubar
parents:
diff changeset
828 nsubj = length(subjects)
0de946608423 Uploaded
fubar
parents:
diff changeset
829 TCols = as.numeric(strsplit(TreatmentCols,",")[[1]])-1
0de946608423 Uploaded
fubar
parents:
diff changeset
830 CCols = as.numeric(strsplit(ControlCols,",")[[1]])-1
0de946608423 Uploaded
fubar
parents:
diff changeset
831 cat('Got TCols=')
0de946608423 Uploaded
fubar
parents:
diff changeset
832 cat(TCols)
0de946608423 Uploaded
fubar
parents:
diff changeset
833 cat('; CCols=')
0de946608423 Uploaded
fubar
parents:
diff changeset
834 cat(CCols)
0de946608423 Uploaded
fubar
parents:
diff changeset
835 cat('\n')
0de946608423 Uploaded
fubar
parents:
diff changeset
836 <![CDATA[
0de946608423 Uploaded
fubar
parents:
diff changeset
837 useCols = c(TCols,CCols)
0de946608423 Uploaded
fubar
parents:
diff changeset
838 if (file.exists(Out_Dir) == F) dir.create(Out_Dir)
0de946608423 Uploaded
fubar
parents:
diff changeset
839 Count_Matrix = read.table(Input,header=T,row.names=1,sep='\t')
0de946608423 Uploaded
fubar
parents:
diff changeset
840 snames = colnames(Count_Matrix)
0de946608423 Uploaded
fubar
parents:
diff changeset
841 nsamples = length(snames)
0de946608423 Uploaded
fubar
parents:
diff changeset
842 if (nsubj > 0 & nsubj != nsamples) {
0de946608423 Uploaded
fubar
parents:
diff changeset
843 options("show.error.messages"=T)
0de946608423 Uploaded
fubar
parents:
diff changeset
844 mess = paste('Fatal error: Supplied subject id list',paste(subjects,collapse=','),
0de946608423 Uploaded
fubar
parents:
diff changeset
845 'has length',nsubj,'but there are',nsamples,'samples',paste(snames,collapse=','))
0de946608423 Uploaded
fubar
parents:
diff changeset
846 write(mess, stderr())
0de946608423 Uploaded
fubar
parents:
diff changeset
847 quit(save="no",status=4)
0de946608423 Uploaded
fubar
parents:
diff changeset
848 }
0de946608423 Uploaded
fubar
parents:
diff changeset
849 if (length(subjects) != 0) {subjects = subjects[useCols]}
0de946608423 Uploaded
fubar
parents:
diff changeset
850 Count_Matrix = Count_Matrix[,useCols] ### reorder columns
0de946608423 Uploaded
fubar
parents:
diff changeset
851 rn = rownames(Count_Matrix)
0de946608423 Uploaded
fubar
parents:
diff changeset
852 islib = rn %in% c('librarySize','NotInBedRegions')
0de946608423 Uploaded
fubar
parents:
diff changeset
853 LibSizes = Count_Matrix[subset(rn,islib),][1] # take first
0de946608423 Uploaded
fubar
parents:
diff changeset
854 Count_Matrix = Count_Matrix[subset(rn,! islib),]
0de946608423 Uploaded
fubar
parents:
diff changeset
855 group = c(rep(TreatmentName,length(TCols)), rep(ControlName,length(CCols)) )
0de946608423 Uploaded
fubar
parents:
diff changeset
856 group = factor(group, levels=c(ControlName,TreatmentName))
0de946608423 Uploaded
fubar
parents:
diff changeset
857 colnames(Count_Matrix) = paste(group,colnames(Count_Matrix),sep="_")
0de946608423 Uploaded
fubar
parents:
diff changeset
858 results = edgeIt(Count_Matrix=Count_Matrix,group=group, out_edgeR=out_edgeR, out_Voom=out_Voom, out_DESeq2=out_DESeq2,
0de946608423 Uploaded
fubar
parents:
diff changeset
859 fdrtype='BH',mydesign=NULL,priordf=edgeR_priordf,fdrthresh=fdrthresh,outputdir='.',
0de946608423 Uploaded
fubar
parents:
diff changeset
860 myTitle=myTitle,useNDF=F,libSize=c(),filterquantile=fQ,subjects=subjects,TreatmentName=TreatmentName,ControlName=ControlName,
0de946608423 Uploaded
fubar
parents:
diff changeset
861 doDESeq2=doDESeq2,doVoom=doVoom,doCamera=doCamera,doedgeR=doedgeR,org=org,
0de946608423 Uploaded
fubar
parents:
diff changeset
862 histgmt=history_gmt,bigmt=builtin_gmt,DESeq_fitType=DESeq_fitType,robust_meth=edgeR_robust_meth)
0de946608423 Uploaded
fubar
parents:
diff changeset
863 sessionInfo()
0de946608423 Uploaded
fubar
parents:
diff changeset
864
0de946608423 Uploaded
fubar
parents:
diff changeset
865 sink()
0de946608423 Uploaded
fubar
parents:
diff changeset
866 ]]>
0de946608423 Uploaded
fubar
parents:
diff changeset
867 </configfile>
143
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
868 </configfiles>
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
869 <help>
132
0de946608423 Uploaded
fubar
parents:
diff changeset
870
0de946608423 Uploaded
fubar
parents:
diff changeset
871 **What it does**
0de946608423 Uploaded
fubar
parents:
diff changeset
872
0de946608423 Uploaded
fubar
parents:
diff changeset
873 Allows short read sequence counts from controlled experiments to be analysed for differentially expressed genes.
0de946608423 Uploaded
fubar
parents:
diff changeset
874 Optionally adds a term for subject if not all samples are independent or if some other factor needs to be blocked in the design.
0de946608423 Uploaded
fubar
parents:
diff changeset
875
0de946608423 Uploaded
fubar
parents:
diff changeset
876 **Input**
0de946608423 Uploaded
fubar
parents:
diff changeset
877
0de946608423 Uploaded
fubar
parents:
diff changeset
878 Requires a count matrix as a tabular file. These are best made using the companion HTSeq_ based counter Galaxy wrapper
0de946608423 Uploaded
fubar
parents:
diff changeset
879 and your fave gene model to generate inputs. Each row is a genomic feature (gene or exon eg) and each column the
0de946608423 Uploaded
fubar
parents:
diff changeset
880 non-negative integer count of reads from one sample overlapping the feature.
0de946608423 Uploaded
fubar
parents:
diff changeset
881
0de946608423 Uploaded
fubar
parents:
diff changeset
882 The matrix must have a header row uniquely identifying the source samples, and unique row names in
0de946608423 Uploaded
fubar
parents:
diff changeset
883 the first column. Typically the row names are gene symbols or probe ids for downstream use in GSEA and other methods.
0de946608423 Uploaded
fubar
parents:
diff changeset
884 They must be unique and R names or they will be mangled - please read the fine R docs for the rules on identifiers.
0de946608423 Uploaded
fubar
parents:
diff changeset
885
0de946608423 Uploaded
fubar
parents:
diff changeset
886 **Specifying comparisons**
0de946608423 Uploaded
fubar
parents:
diff changeset
887
0de946608423 Uploaded
fubar
parents:
diff changeset
888 This is basically dumbed down for two factors - case vs control.
0de946608423 Uploaded
fubar
parents:
diff changeset
889
0de946608423 Uploaded
fubar
parents:
diff changeset
890 More complex interfaces are possible but painful at present.
0de946608423 Uploaded
fubar
parents:
diff changeset
891 Probably need to specify a phenotype file to do this better.
0de946608423 Uploaded
fubar
parents:
diff changeset
892 Work in progress. Send code.
0de946608423 Uploaded
fubar
parents:
diff changeset
893
0de946608423 Uploaded
fubar
parents:
diff changeset
894 If you have (eg) paired samples and wish to include a term in the GLM to account for some other factor (subject in the case of paired samples),
0de946608423 Uploaded
fubar
parents:
diff changeset
895 put a comma separated list of indicators for every sample (whether modelled or not!) indicating (eg) the subject number or
0de946608423 Uploaded
fubar
parents:
diff changeset
896 A list of integers, one for each subject or an empty string if samples are all independent.
0de946608423 Uploaded
fubar
parents:
diff changeset
897 If not empty, there must be exactly as many integers in the supplied integer list as there are columns (samples) in the count matrix.
0de946608423 Uploaded
fubar
parents:
diff changeset
898 Integers for samples that are not in the analysis *must* be present in the string as filler even if not used.
0de946608423 Uploaded
fubar
parents:
diff changeset
899
0de946608423 Uploaded
fubar
parents:
diff changeset
900 So if you have 2 pairs out of 6 samples, you need to put in unique integers for the unpaired ones
0de946608423 Uploaded
fubar
parents:
diff changeset
901 eg if you had 6 samples with the first two independent but the second and third pairs each being from independent subjects. you might use
0de946608423 Uploaded
fubar
parents:
diff changeset
902 8,9,1,1,2,2
0de946608423 Uploaded
fubar
parents:
diff changeset
903 as subject IDs to indicate two paired samples from the same subject in columns 3/4 and 5/6
0de946608423 Uploaded
fubar
parents:
diff changeset
904
0de946608423 Uploaded
fubar
parents:
diff changeset
905 **Methods available**
0de946608423 Uploaded
fubar
parents:
diff changeset
906
0de946608423 Uploaded
fubar
parents:
diff changeset
907 You can run 3 popular Bioconductor packages available for count data.
0de946608423 Uploaded
fubar
parents:
diff changeset
908
0de946608423 Uploaded
fubar
parents:
diff changeset
909 edgeR - see edgeR_ for details
0de946608423 Uploaded
fubar
parents:
diff changeset
910
0de946608423 Uploaded
fubar
parents:
diff changeset
911 VOOM/limma - see limma_VOOM_ for details
0de946608423 Uploaded
fubar
parents:
diff changeset
912
0de946608423 Uploaded
fubar
parents:
diff changeset
913 DESeq2 - see DESeq2_ for details
0de946608423 Uploaded
fubar
parents:
diff changeset
914
0de946608423 Uploaded
fubar
parents:
diff changeset
915 and optionally camera in edgeR which works better if MSigDB is installed.
0de946608423 Uploaded
fubar
parents:
diff changeset
916
0de946608423 Uploaded
fubar
parents:
diff changeset
917 **Outputs**
0de946608423 Uploaded
fubar
parents:
diff changeset
918
0de946608423 Uploaded
fubar
parents:
diff changeset
919 Some helpful plots and analysis results. Note that most of these are produced using R code
0de946608423 Uploaded
fubar
parents:
diff changeset
920 suggested by the excellent documentation and vignettes for the Bioconductor
0de946608423 Uploaded
fubar
parents:
diff changeset
921 packages invoked. The Tool Factory is used to automatically lay these out for you to enjoy.
0de946608423 Uploaded
fubar
parents:
diff changeset
922
0de946608423 Uploaded
fubar
parents:
diff changeset
923 **Note on Voom**
0de946608423 Uploaded
fubar
parents:
diff changeset
924
0de946608423 Uploaded
fubar
parents:
diff changeset
925 The voom from limma version 3.16.6 help in R includes this from the authors - but you should read the paper to interpret this method.
0de946608423 Uploaded
fubar
parents:
diff changeset
926
0de946608423 Uploaded
fubar
parents:
diff changeset
927 This function is intended to process RNA-Seq or ChIP-Seq data prior to linear modelling in limma.
0de946608423 Uploaded
fubar
parents:
diff changeset
928
0de946608423 Uploaded
fubar
parents:
diff changeset
929 voom is an acronym for mean-variance modelling at the observational level.
0de946608423 Uploaded
fubar
parents:
diff changeset
930 The key concern is to estimate the mean-variance relationship in the data, then use this to compute appropriate weights for each observation.
0de946608423 Uploaded
fubar
parents:
diff changeset
931 Count data almost show non-trivial mean-variance relationships. Raw counts show increasing variance with increasing count size, while log-counts typically show a decreasing mean-variance trend.
0de946608423 Uploaded
fubar
parents:
diff changeset
932 This function estimates the mean-variance trend for log-counts, then assigns a weight to each observation based on its predicted variance.
0de946608423 Uploaded
fubar
parents:
diff changeset
933 The weights are then used in the linear modelling process to adjust for heteroscedasticity.
0de946608423 Uploaded
fubar
parents:
diff changeset
934
0de946608423 Uploaded
fubar
parents:
diff changeset
935 In an experiment, a count value is observed for each tag in each sample. A tag-wise mean-variance trend is computed using lowess.
0de946608423 Uploaded
fubar
parents:
diff changeset
936 The tag-wise mean is the mean log2 count with an offset of 0.5, across samples for a given tag.
0de946608423 Uploaded
fubar
parents:
diff changeset
937 The tag-wise variance is the quarter-root-variance of normalized log2 counts per million values with an offset of 0.5, across samples for a given tag.
0de946608423 Uploaded
fubar
parents:
diff changeset
938 Tags with zero counts across all samples are not included in the lowess fit. Optional normalization is performed using normalizeBetweenArrays.
0de946608423 Uploaded
fubar
parents:
diff changeset
939 Using fitted values of log2 counts from a linear model fit by lmFit, variances from the mean-variance trend were interpolated for each observation.
0de946608423 Uploaded
fubar
parents:
diff changeset
940 This was carried out by approxfun. Inverse variance weights can be used to correct for mean-variance trend in the count data.
0de946608423 Uploaded
fubar
parents:
diff changeset
941
0de946608423 Uploaded
fubar
parents:
diff changeset
942
0de946608423 Uploaded
fubar
parents:
diff changeset
943 Author(s)
0de946608423 Uploaded
fubar
parents:
diff changeset
944
0de946608423 Uploaded
fubar
parents:
diff changeset
945 Charity Law and Gordon Smyth
0de946608423 Uploaded
fubar
parents:
diff changeset
946
0de946608423 Uploaded
fubar
parents:
diff changeset
947 References
0de946608423 Uploaded
fubar
parents:
diff changeset
948
0de946608423 Uploaded
fubar
parents:
diff changeset
949 Law, CW (2013). Precision weights for gene expression analysis. PhD Thesis. University of Melbourne, Australia.
0de946608423 Uploaded
fubar
parents:
diff changeset
950
0de946608423 Uploaded
fubar
parents:
diff changeset
951 Law, CW, Chen, Y, Shi, W, Smyth, GK (2013). Voom! Precision weights unlock linear model analysis tools for RNA-seq read counts.
0de946608423 Uploaded
fubar
parents:
diff changeset
952 Technical Report 1 May 2013, Bioinformatics Division, Walter and Eliza Hall Institute of Medical Reseach, Melbourne, Australia.
0de946608423 Uploaded
fubar
parents:
diff changeset
953 http://www.statsci.org/smyth/pubs/VoomPreprint.pdf
0de946608423 Uploaded
fubar
parents:
diff changeset
954
0de946608423 Uploaded
fubar
parents:
diff changeset
955 See Also
0de946608423 Uploaded
fubar
parents:
diff changeset
956
0de946608423 Uploaded
fubar
parents:
diff changeset
957 A voom case study is given in the edgeR User's Guide.
0de946608423 Uploaded
fubar
parents:
diff changeset
958
0de946608423 Uploaded
fubar
parents:
diff changeset
959 vooma is a similar function but for microarrays instead of RNA-seq.
0de946608423 Uploaded
fubar
parents:
diff changeset
960
0de946608423 Uploaded
fubar
parents:
diff changeset
961
0de946608423 Uploaded
fubar
parents:
diff changeset
962 ***old rant on changes to Bioconductor package variable names between versions***
0de946608423 Uploaded
fubar
parents:
diff changeset
963
0de946608423 Uploaded
fubar
parents:
diff changeset
964 The edgeR authors made a small cosmetic change in the name of one important variable (from p.value to PValue)
0de946608423 Uploaded
fubar
parents:
diff changeset
965 breaking this and all other code that assumed the old name for this variable,
0de946608423 Uploaded
fubar
parents:
diff changeset
966 between edgeR2.4.4 and 2.4.6 (the version for R 2.14 as at the time of writing).
0de946608423 Uploaded
fubar
parents:
diff changeset
967 This means that all code using edgeR is sensitive to the version. I think this was a very unwise thing
0de946608423 Uploaded
fubar
parents:
diff changeset
968 to do because it wasted hours of my time to track down and will similarly cost other edgeR users dearly
0de946608423 Uploaded
fubar
parents:
diff changeset
969 when their old scripts break. This tool currently now works with 2.4.6.
0de946608423 Uploaded
fubar
parents:
diff changeset
970
0de946608423 Uploaded
fubar
parents:
diff changeset
971 **Note on prior.N**
0de946608423 Uploaded
fubar
parents:
diff changeset
972
0de946608423 Uploaded
fubar
parents:
diff changeset
973 http://seqanswers.com/forums/showthread.php?t=5591 says:
0de946608423 Uploaded
fubar
parents:
diff changeset
974
0de946608423 Uploaded
fubar
parents:
diff changeset
975 *prior.n*
0de946608423 Uploaded
fubar
parents:
diff changeset
976
0de946608423 Uploaded
fubar
parents:
diff changeset
977 The value for prior.n determines the amount of smoothing of tagwise dispersions towards the common dispersion.
0de946608423 Uploaded
fubar
parents:
diff changeset
978 You can think of it as like a "weight" for the common value. (It is actually the weight for the common likelihood
0de946608423 Uploaded
fubar
parents:
diff changeset
979 in the weighted likelihood equation). The larger the value for prior.n, the more smoothing, i.e. the closer your
0de946608423 Uploaded
fubar
parents:
diff changeset
980 tagwise dispersion estimates will be to the common dispersion. If you use a prior.n of 1, then that gives the
0de946608423 Uploaded
fubar
parents:
diff changeset
981 common likelihood the weight of one observation.
0de946608423 Uploaded
fubar
parents:
diff changeset
982
0de946608423 Uploaded
fubar
parents:
diff changeset
983 In answer to your question, it is a good thing to squeeze the tagwise dispersions towards a common value,
0de946608423 Uploaded
fubar
parents:
diff changeset
984 or else you will be using very unreliable estimates of the dispersion. I would not recommend using the value that
0de946608423 Uploaded
fubar
parents:
diff changeset
985 you obtained from estimateSmoothing()---this is far too small and would result in virtually no moderation
0de946608423 Uploaded
fubar
parents:
diff changeset
986 (squeezing) of the tagwise dispersions. How many samples do you have in your experiment?
0de946608423 Uploaded
fubar
parents:
diff changeset
987 What is the experimental design? If you have few samples (less than 6) then I would suggest a prior.n of at least 10.
0de946608423 Uploaded
fubar
parents:
diff changeset
988 If you have more samples, then the tagwise dispersion estimates will be more reliable,
0de946608423 Uploaded
fubar
parents:
diff changeset
989 so you could consider using a smaller prior.n, although I would hesitate to use a prior.n less than 5.
0de946608423 Uploaded
fubar
parents:
diff changeset
990
0de946608423 Uploaded
fubar
parents:
diff changeset
991
0de946608423 Uploaded
fubar
parents:
diff changeset
992 From Bioconductor Digest, Vol 118, Issue 5, Gordon writes:
0de946608423 Uploaded
fubar
parents:
diff changeset
993
0de946608423 Uploaded
fubar
parents:
diff changeset
994 Dear Dorota,
0de946608423 Uploaded
fubar
parents:
diff changeset
995
0de946608423 Uploaded
fubar
parents:
diff changeset
996 The important settings are prior.df and trend.
0de946608423 Uploaded
fubar
parents:
diff changeset
997
0de946608423 Uploaded
fubar
parents:
diff changeset
998 prior.n and prior.df are related through prior.df = prior.n * residual.df,
0de946608423 Uploaded
fubar
parents:
diff changeset
999 and your experiment has residual.df = 36 - 12 = 24. So the old setting of
0de946608423 Uploaded
fubar
parents:
diff changeset
1000 prior.n=10 is equivalent for your data to prior.df = 240, a very large
0de946608423 Uploaded
fubar
parents:
diff changeset
1001 value. Going the other way, the new setting of prior.df=10 is equivalent
0de946608423 Uploaded
fubar
parents:
diff changeset
1002 to prior.n=10/24.
0de946608423 Uploaded
fubar
parents:
diff changeset
1003
0de946608423 Uploaded
fubar
parents:
diff changeset
1004 To recover old results with the current software you would use
0de946608423 Uploaded
fubar
parents:
diff changeset
1005
0de946608423 Uploaded
fubar
parents:
diff changeset
1006 estimateTagwiseDisp(object, prior.df=240, trend="none")
0de946608423 Uploaded
fubar
parents:
diff changeset
1007
0de946608423 Uploaded
fubar
parents:
diff changeset
1008 To get the new default from old software you would use
0de946608423 Uploaded
fubar
parents:
diff changeset
1009
0de946608423 Uploaded
fubar
parents:
diff changeset
1010 estimateTagwiseDisp(object, prior.n=10/24, trend=TRUE)
0de946608423 Uploaded
fubar
parents:
diff changeset
1011
0de946608423 Uploaded
fubar
parents:
diff changeset
1012 Actually the old trend method is equivalent to trend="loess" in the new
0de946608423 Uploaded
fubar
parents:
diff changeset
1013 software. You should use plotBCV(object) to see whether a trend is
0de946608423 Uploaded
fubar
parents:
diff changeset
1014 required.
0de946608423 Uploaded
fubar
parents:
diff changeset
1015
0de946608423 Uploaded
fubar
parents:
diff changeset
1016 Note you could also use
0de946608423 Uploaded
fubar
parents:
diff changeset
1017
0de946608423 Uploaded
fubar
parents:
diff changeset
1018 prior.n = getPriorN(object, prior.df=10)
0de946608423 Uploaded
fubar
parents:
diff changeset
1019
0de946608423 Uploaded
fubar
parents:
diff changeset
1020 to map between prior.df and prior.n.
0de946608423 Uploaded
fubar
parents:
diff changeset
1021
0de946608423 Uploaded
fubar
parents:
diff changeset
1022 ----
0de946608423 Uploaded
fubar
parents:
diff changeset
1023
0de946608423 Uploaded
fubar
parents:
diff changeset
1024 **Attributions**
0de946608423 Uploaded
fubar
parents:
diff changeset
1025
0de946608423 Uploaded
fubar
parents:
diff changeset
1026 edgeR - edgeR_
0de946608423 Uploaded
fubar
parents:
diff changeset
1027
0de946608423 Uploaded
fubar
parents:
diff changeset
1028 VOOM/limma - limma_VOOM_
0de946608423 Uploaded
fubar
parents:
diff changeset
1029
0de946608423 Uploaded
fubar
parents:
diff changeset
1030 DESeq2 - DESeq2_ for details
0de946608423 Uploaded
fubar
parents:
diff changeset
1031
0de946608423 Uploaded
fubar
parents:
diff changeset
1032 See above for Bioconductor package documentation for packages exposed in Galaxy by this tool and app store package.
0de946608423 Uploaded
fubar
parents:
diff changeset
1033
0de946608423 Uploaded
fubar
parents:
diff changeset
1034 Galaxy_ (that's what you are using right now!) for gluing everything together
0de946608423 Uploaded
fubar
parents:
diff changeset
1035
0de946608423 Uploaded
fubar
parents:
diff changeset
1036 Otherwise, all code and documentation comprising this tool was written by Ross Lazarus and is
0de946608423 Uploaded
fubar
parents:
diff changeset
1037 licensed to you under the LGPL_ like other rgenetics artefacts
0de946608423 Uploaded
fubar
parents:
diff changeset
1038
0de946608423 Uploaded
fubar
parents:
diff changeset
1039 .. _LGPL: http://www.gnu.org/copyleft/lesser.html
0de946608423 Uploaded
fubar
parents:
diff changeset
1040 .. _HTSeq: http://www-huber.embl.de/users/anders/HTSeq/doc/index.html
0de946608423 Uploaded
fubar
parents:
diff changeset
1041 .. _edgeR: http://www.bioconductor.org/packages/release/bioc/html/edgeR.html
0de946608423 Uploaded
fubar
parents:
diff changeset
1042 .. _DESeq2: http://www.bioconductor.org/packages/release/bioc/html/DESeq2.html
0de946608423 Uploaded
fubar
parents:
diff changeset
1043 .. _limma_VOOM: http://www.bioconductor.org/packages/release/bioc/html/limma.html
0de946608423 Uploaded
fubar
parents:
diff changeset
1044 .. _Galaxy: http://getgalaxy.org
0de946608423 Uploaded
fubar
parents:
diff changeset
1045 </help>
143
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
1046 <citations>
142
e7894f37320a Uploaded
iuc
parents: 140
diff changeset
1047 <citation type="doi">doi: 10.1093/bioinformatics/btp616</citation>
143
1435811cbf01 Uploaded
iuc
parents: 142
diff changeset
1048 </citations>
132
0de946608423 Uploaded
fubar
parents:
diff changeset
1049 </tool>