pam50.xml

<tool id='pam50' name="PAM50" version='1.0.0.1'>
    <description>applies the PAM50 algorithm</description>
    <requirements>
        <requirement type="package" version="3.5.1">r</requirement>
    </requirements>
    <command detect_errors="exit_code"><![CDATA[
    Rscript "$script"
]]>
    </command>
    <configfiles>
        <configfile name="script">

centroids_file = "$__tool_directory__/tool-data/pam50_centroids_update.txt"
input = "$input"
method = "$method"

## Load centroid file
centroids = read.csv(centroids_file, check.names = FALSE, sep="\t", row.names = 1)

## Load gene expression file (columns samples, rows genes)
exp = read.csv(input, sep="\t", row.names = 1, header=TRUE, check.names = FALSE)

## Log transform matrix
# log2(x+1) the ruv fixed matrix
l.exp=log2(exp+1)

## Z-SCALING: Calculate z-score for each gene across all samples: z-score=(value-gene mean)/gene sd
sctl.exp = t(scale(data.frame(t(l.exp), check.names =F), scale=T, center=TRUE))

# Remove rows were sd=0 in z-score calculation
sctl.exp = sctl.exp[complete.cases(sctl.exp),]

# Convert to Matrix
sctl.exp.mat = as.matrix((sctl.exp))

## CORRELATION WITH CENTROIDS
scor=round(cor(sctl.exp.mat[intersect(rownames(sctl.exp.mat), rownames(centroids)),], centroids[intersect(rownames(sctl.exp.mat), rownames(centroids)),], method=method),2)

#write to output
write.table(exp, file="gene_exp.tsv", row.names=T,  sep="\t", col.names=NA, quote=F)
write.table(scor, file="scor.tsv", row.names=T, sep="\t", col.names=NA, quote=F)

        </configfile>
    </configfiles>
    <inputs>
        <param type="data" name="input" label="Sample Matrix" format="tsv,tabular,txt" help="All samples in matrix should be breast cancers for most accurate prediction of subtype."/>
        <param type="select" name="method" label="Correlation method" help="Type of correlation to perform">
            <option value="pearson">Pearson</option>
            <option value="spearman">Spearman</option>
        </param>
    </inputs>
    <outputs>
        <data format="tsv" name="output" label="${tool.name} on $on_string: PAM50 CORRELATION" from_work_dir="scor.tsv"/>
        <data format="tsv" name="output2" label="${tool.name} on $on_string: EXPRESSION" from_work_dir="gene_exp.tsv"/>
    </outputs>
</tool>