compare.analysis: Exporting Hit Candidate Gene Information

In caRpools: CRISPR AnalyzeR for Pooled CRISPR Screens

Description

Although the candidate lists of each analysis method can be saved separately, caRpools offer a comparative approach, which creates tables that include the information from all analysis methods at once for a faster overview.

This is done using the function 'compare.analysis', which offers not only ouptu for Venn Diagrams, but also for tables.

Usage

compare.analysis(wilcox=NULL, deseq=NULL, mageck=NULL, type="enriched",
cutoff.deseq = NULL, cutoff.wilcox = NULL, cutoff.mageck = NULL,
cutoff.override=TRUE, cutoff.hits=5, output="list",
sort.by=c("mageck","pval","fdr"), plot.method=c("wilcox","mageck", "deseq"),
plot.feature=c("pval","fdr","pval"), pch=16)

Arguments

wilcox	Data output from 'stat.wilcox'. *Default* NULL *Values* Data output from 'stat.wilcox'.
deseq	Data output from 'stat.deseq'. *Default* NULL *Values* Data output from 'stat.deseq'.
mageck	Data output from 'stat.mageck'. *Default* NULL *Values* Data output from 'stat.mageck'.
type	Either enriched or depleted.
cutoff.deseq	P-Value threshold used to determine significance. *Default* 0.001 *Values* numeric
cutoff.wilcox	P-Value threshold used to determine significance. *Default* 0.001 *Values* numeric
cutoff.mageck	P-Value threshold used to determine significance. *Default* 0.001 *Values* numeric
cutoff.override	Shall the p-value threshold be ignored? If this is TRUE, the top percentage gene of 'cutoff.hits' is used instead. *Default* FALSE *Values* TRUE, FALSE
cutoff.hits	The percentatge of top genes being used if 'cutoff.override=TRUE'. *Default** NULL *Values* numeric
output	Three different types of output can be generated: A list with all genes including the information from 'stat.wilcox', 'stat.DEseq' and 'stat.mageck', a sorted ranked output, a venn diagram compatible output and 3D scatterplot. *Default* "list" *Values* "list", "rank", "venn", "3dplot"
sort.by	This indicates the sorting for 'type="rank" and type="list"' and is a vector. By default, data is sorted by the FDR of MAGeCK. needs to be a vector. *Default* c("mageck","fdr","fdr") *Values* c("mageck","fdr","fdr"), c("mageck","fdr","rank"), c("mageck","fdr","rank"), c("wilcox","pval","pval"), c("wilcox","pval","genes"), c("deseq","pval","pval"), c("deseq","pval","genes")
plot.method	Used only if 'type="3dplot"'. This indicates what is plotted at the x, y and z-axis and thus needs to be a vector of length 3. *Default* c("wilcox","mageck", "deseq"), plots wilcox on X-axis, mageck on y-axis and deseq on z-axis *Values* c("wilcox","mageck", "deseq") or any other combination
plot.feature	If 'type="3dplot"', this indicates the type of data plotted on each axis of the 3d plot. This can only be set according to the features available of the method used to be plotted as indicated in 'plot.method'. *Default* c("pval","fdr", "pval") which uses the p-value of wilcox, the fdr or MAGeCK and p-value of DESeq2. *Values* c("pval","fdr", "pval"), or ANY combination according to 'plot.method'
pch	The type of point used in the plot. See '?par()'. *Default* 16 *Values* Any number describing the point, e.g. 16 (numeric)

Details

none

Value

Returns a table with information.

Note

none

Author(s)

Jan Winter

Examples

data(caRpools)

data.wilcox = stat.wilcox(untreated.list = list(CONTROL1, CONTROL2),
  treated.list = list(TREAT1,TREAT2), namecolumn=1, fullmatchcolumn=2,
  normalize=TRUE, norm.fun=median, sorting=FALSE, controls="random",
  control.picks=NULL)
  
data.deseq = stat.DESeq(untreated.list = list(CONTROL1, CONTROL2),
  treated.list = list(TREAT1,TREAT2), namecolumn=1,
  fullmatchcolumn=2, extractpattern=expression("^(.+?)(_.+)"),
  sorting=FALSE, filename.deseq = "ANALYSIS-DESeq2-sgRNA.tab",
  fitType="parametric")
  
data.mageck = stat.mageck(untreated.list = list(CONTROL1, CONTROL2),
treated.list = list(TREAT1,TREAT2), namecolumn=1, fullmatchcolumn=2,
norm.fun="median", extractpattern=expression("^(.+?)(_.+)"),
mageckfolder=NULL, sort.criteria="neg", adjust.method="fdr",
filename = "TEST" , fdr.pval = 0.05)

# Perform the comparison
data.analysis.enriched = compare.analysis(wilcox=data.wilcox,
    deseq=data.deseq, mageck=data.mageck, type="enriched",
    cutoff.override = FALSE, cutoff.hits=NULL, output="list",
    sort.by=c("mageck","fdr","rank"))
## Write to a file
xlsx::write.xlsx(data.analysis.enriched,
    file="COMPARE-HITS.xls",
    sheetName="Enriched")
# Print to console
knitr::kable(data.analysis.enriched[1:10,c(2:7)])

caRpools documentation built on May 2, 2019, 11:26 a.m.