Mercurial > repos > mingchen0919 > rmarkdown_deseq2_count_matrix

--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/rmarkdown_deseq2_count_matrix.Rmd	Sat Dec 30 00:25:38 2017 -0500
@@ -0,0 +1,39 @@
+---
+title: 'DESeq2 analysis'
+output:
+    html_document:
+      number_sections: true
+      toc: true
+      theme: cosmo
+      highlight: tango
+---
+
+```{r setup, include=FALSE, warning=FALSE, message=FALSE}
+knitr::opts_chunk$set(
+  echo = opt$echo,
+  error = TRUE
+)
+```
+
+
+## User input
+
+```{r 'user input'}
+df = data.frame(name = names(opt)[-1],
+                value = unlist(opt))
+df
+```
+
+
+## Count Matrix
+
+```{r 'count matrix'}
+count_data = read.table(opt$count_data)
+head(count_data, 10)
+```
+
+```{r 'ste[ 2'}
+
+```
+
+
--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/rmarkdown_deseq2_count_matrix.xml	Sat Dec 30 00:25:38 2017 -0500
@@ -0,0 +1,73 @@
+<tool id="rmarkdown_deseq2_count_matrix" name="DESeq2 (count matrix)" version="1.0.0">
+    <requirements>
+        <requirement type="package" version="1.15.0.6-0">pandoc</requirement>
+        <requirement type="package" version="1.20.0">r-getopt</requirement>
+        <requirement type="package" version="1.3">r-rmarkdown</requirement>
+        <requirement type="package" version="0.3.5">r-htmltools</requirement>
+        <requirement type="package" version="0.5.0">r-dplyr</requirement>
+        <requirement type="package" version="1.14.1">bioconductor-deseq2</requirement>
+        <requirement type="package" version="0.2">r-dt</requirement>
+    </requirements>
+    <description>deseq2 analysis given a count matrix input</description>
+    <stdio>
+        <regex source="stderr" match="XXX" level="warning"
+               description="Check the warnings_and_errors.txt file for more details."/>
+    </stdio>
+    <command><![CDATA[
+Rscript '${__tool_directory__}/rmarkdown_deseq2_count_matrix_render.R'
+                        -e $echo
+                        -c $count_data
+                        -C $col_data
+                        -D '$design_formula'
+                        -k $contrast_condition
+                        -T $treatment
+                        -k $control
+                        -r $report
+                        -d $report.files_path
+                        -s $sink_message
+                        -t '${__tool_directory__}/rmarkdown_deseq2_count_matrix.Rmd'
+                    ]]></command>
+    <inputs>
+        <param type="boolean" name="echo" label="Display analysis code in report?" optional="False" checked="False"
+               truevalue="TRUE" falsevalue="FALSE"/>
+        <param type="data" name="count_data" label="count data" optional="False" format="txt,csv" multiple="False"/>
+        <param type="data" name="col_data" label="column data" optional="False" format="txt,csv" multiple="False"/>
+        <param type="text" name="design_formula" label="design formula"
+               help="The simplest design formula for differential expression would be ~ condition, where condition&#13;&#10;                     is a column in colData(dds) that specifies which of two (or more groups) the samples belong to"
+               optional="False" value="~ condition_1 + condition_2"/>
+        <param type="text" name="contrast_condition" label="condition"
+               help="A condition name specified in the design formula" optional="False" value="condition_1"/>
+        <param type="text" name="treatment" label="treatment" help="the name of the numerator level for the fold change"
+               optional="False"/>
+        <param type="text" name="control" label="control" help="the name of the denominator level for the fold change"
+               optional="False"/>
+    </inputs>
+    <outputs>
+        <data name="report" format="html" label="DESeq2 report" hidden="false"/>
+        <data name="differential_genes" format="csv" label="differential genes" from_work_dir="differential_genes.csv"
+              hidden="false"/>
+        <data name="sink_message" format="txt" label="Warnings and Errors" from_work_dir="warnings_and_errors.txt"
+              hidden="false"/>
+    </outputs>
+    <help><![CDATA[See http://bioconductor.org/packages/devel/bioc/vignettes/DESeq2/inst/doc/DESeq2.html]]></help>
+    <citations>
+        <citation type="bibtex"><![CDATA[            @article{love2014moderated,
+            title={Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2},
+            author={Love, Michael I and Huber, Wolfgang and Anders, Simon},
+            journal={Genome biology},
+            volume={15},
+            number={12},
+            pages={550},
+            year={2014},
+            publisher={BioMed Central}
+            }]]></citation>
+        <citation type="bibtex"><![CDATA[            @article{allaire2016rmarkdown,
+            title={rmarkdown: Dynamic Documents for R, 2016},
+            author={Allaire, J and Cheng, Joe and Xie, Yihui and McPherson, Jonathan and Chang, Winston and Allen, Jeff
+            and Wickham, Hadley and Atkins, Aron and Hyndman, Rob},
+            journal={R package version 0.9},
+            volume={6},
+            year={2016}
+            }]]></citation>
+    </citations>
+</tool>
--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/rmarkdown_deseq2_count_matrix_render.R	Sat Dec 30 00:25:38 2017 -0500
@@ -0,0 +1,83 @@
+library(getopt)
+library(rmarkdown)
+library(htmltools)
+library(dplyr)
+
+##============ Sink warnings and errors to a file ==============
+## use the sink() function to wrap all code within it.
+##==============================================================
+zz = file('warnings_and_errors.txt')
+sink(zz)
+sink(zz, type = 'message')
+  ##---------below is the code for rendering .Rmd templates-----
+
+  ##=============STEP 1: handle command line arguments==========
+  ##
+  ##============================================================
+  # column 1: the long flag name
+  # column 2: the short flag alias. A SINGLE character string
+  # column 3: argument mask
+  #           0: no argument
+  #           1: argument required
+  #           2: argument is optional
+  # column 4: date type to which the flag's argument shall be cast.
+  #           possible values: logical, integer, double, complex, character.
+  #-------------------------------------------------------------
+  #++++++++++++++++++++ Best practice ++++++++++++++++++++++++++
+  # 1. short flag alias should match the flag in the command section in the XML file.
+  # 2. long flag name can be any legal R variable names
+  # 3. two names in args_list can have common string but one name should not be a part of another name.
+  #    for example, one name is "ECHO", if another name is "ECHO_XXX", it will cause problems.
+  #+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
+  args_list=list()
+  ##------- 1. input data ---------------------
+  args_list$ECHO = c('echo', 'e', '1', 'character')
+  args_list$c = c('count_data', 'c', '1', 'character')
+  args_list$C = c('col_data', 'C', '1', 'character')
+  args_list$D = c('design_formula', 'D', '1', 'character')
+  args_list$k = c('contrast_condition', 'k', '1', 'character')
+  args_list$T = c('treatment', 'T', '1', 'character')
+  args_list$K = c('control', 'K', '1', 'character')
+  ##--------2. output report and outputs --------------
+  args_list$REPORT_HTML = c('report_html', 'r', '1', 'character')
+  args_list$REPORT_DIR = c('report_dir', 'd', '1', 'character')
+  args_list$SINK_MESSAGE = c('sink_message', 's', '1', 'character')
+  ##--------3. .Rmd templates in the tool directory ----------
+  args_list$TOOL_TEMPLATE_RMD = c('tool_template_rmd', 't', '1', 'character')
+  ##-----------------------------------------------------------
+  opt = getopt(t(as.data.frame(args_list)))
+
+
+
+  ##=======STEP 2: create report directory (optional)==========
+  ##
+  ##===========================================================
+  dir.create(opt$report_dir)
+
+  ##=STEP 3: replace placeholders in .Rmd with argument values=
+  ##
+  ##===========================================================
+  #++ need to replace placeholders with args values one by one+
+  readLines(opt$tool_template_rmd) %>%
+    # (function(x) {
+    #   gsub('ECHO', opt$echo, x)
+    # }) %>%
+    # (function(x) {
+    #   gsub('REPORT_DIR', opt$report_dir, x)
+    # }) %>%
+    (function(x) {
+      fileConn = file('tool_template.Rmd')
+      writeLines(x, con=fileConn)
+      close(fileConn)
+    })
+
+
+  ##=============STEP 4: render .Rmd templates=================
+  ##
+  ##===========================================================
+  render('tool_template.Rmd', output_file = opt$report_html)
+
+
+  ##--------end of code rendering .Rmd templates----------------
+sink()
+##=========== End of sinking output=============================