view deseq2_normalization.xml @ 0:6b990f6c6ba9 draft

planemo upload for repository https://github.com/ARTbio/tools-artbio/tree/master/tools/deseq2_normalization commit 9dfb5a5331b87d213f495febbfccff0f54becb73
author artbio
date Mon, 08 Jan 2018 04:21:24 -0500
parents
children ca845e480c89
line wrap: on
line source

<tool id="deseq2_normalization" name="DESeq2 normalization" version="1.0.0">
	<description>of hit lists (geometric method)</description>
    <requirements>
      <requirement type="package" version="1.8.2gx=r3.2.1_0">bioconductor-deseq2</requirement>
    </requirements>
    <stdio>
        <exit_code range="1:" level="fatal" description="Tool exception" />
    </stdio>
    <command detect_errors="exit_code"><![CDATA[ Rscript $normalizationDESeq ]]></command>
    <configfiles>
        <configfile name="normalizationDESeq">
      ## Setup R error handling to go to stderr
      options( show.error.messages=F,  error = function () { cat( geterrmessage(), file=stderr() ); q( "no", 1, F ) } )
      suppressMessages(require(DESeq2))
      countsTable = read.delim("${input}", header=TRUE, check.names=FALSE)
      rownames( countsTable )= countsTable[,1]
      countsTable= countsTable[ , -1 ]
      conds = factor(c(rep ("C", length(countsTable[1,])-1 ), "T"))
      cds = DESeqDataSetFromMatrix(countsTable, DataFrame(conds), ~ conds)
      cds = estimateSizeFactors( cds )
      write.table (as.data.frame(sizeFactors(cds)), file="${output_factors}", row.names=TRUE, col.names=TRUE, quote= FALSE, dec = ".", sep = "\t", eol = "\n")
      norm_table = as.data.frame(counts( cds, normalized=TRUE ))
      norm_table = cbind(rownames(norm_table), norm_table)
      colnames(norm_table) = c("gene", colnames(countsTable) )
      write.table (norm_table, file="${norm_hit_table}", row.names=FALSE, col.names=TRUE, quote= FALSE, dec = ".", sep = "\t", eol = "\n")
        </configfile>
    </configfiles>
	<inputs>
		<param name="input" type="data" format="tabular" label="Hit lists of items"/>
	</inputs>
	<outputs>
                <data name="output_factors" format="tabular" label="Normalization Factors (Geometric mean method)" />
                <data name="norm_hit_table" format="tabular" label="Normalized Hit Table (Geometric mean method)" />
	</outputs>
    <tests>
        <test>
            <param name="input" value="counts.tab" ftype="tabular"/>
            <output name="output_factors" file="norm_factors.tab" ftype="tabular"/>
            <output name="norm_hit_table" file="norm_table.tab" ftype="tabular"/>
        </test>
    </tests>
    <help>

**What it does**

DESeq2_normalization takes a tabular table of raw read counts (integers), each column corresponding to a sample count,
and returns (1) the DESeq2 Size Factors (geometic mean method) and (2) a normalized table (floatting number)
of counts.

**Inputs**

The following table gives an example of the tabular input that has to be provided to the tool:

<![CDATA[

Geneid	Oocyte 1	Oocyte 2	Pronuclei 1	pronuclei 2

NM_001195025	0	0	0	0

NM_028778	0	0	0	0

NM_001290393	0	0	0	0

NM_008922	3133	2276	2288	2659

NM_175370	4	22	7	16

NM_001290390	0	0	0	0

NM_175642	977	1711	1013	459

NM_001290392	0	0	0	0

NM_027671	0	0	0	0

NM_178884	0	0	0	2

NM_199021	332	413	1436	836

NM_198680	0	0	0	0

NM_011465	0	4	0	0

NM_130890	0	0	0	0

NM_001113393	0	0	0	0

NM_001113392	0	0	0	0

NM_001113391	19	3	88	15

NM_031162	0	0	0	0

NR_103716	0	0	0	0

NM_001160371	0	0	0	0

NM_177834	1	3	2	0

NM_145381	606	445	367	306

]]>


    </help>
    <citations>
        <citation type="doi">10.1186/s13059-014-0550-8</citation>
    </citations>
</tool>