Package 'HD2013SGI'

Title: Mapping genetic interactions in human cancer cells with RNAi and multiparametric phenotyping
Description: This package contains the experimental data and a complete executable transcript (vignette) of the analysis of the HCT116 genetic interaction matrix presented in the paper "Mapping genetic interactions in human cancer cells with RNAi and multiparametric phenotyping" by C. Laufer, B. Fischer, M. Billmann, W. Huber, M. Boutros; Nature Methods (2013) 10:427-31. doi: 10.1038/nmeth.2436.
Authors: Bernd Fischer
Maintainer: Mike Smith <[email protected]>
License: Artistic-2.0
Version: 1.52.0
Built: 2026-10-10 11:06:08 UTC
Source: https://github.com/bioc/HD2013SGI

Help Index

Experimental Data and Analysis of the HCT116 Genetic Interaction Matrix

Description

Experimental Data and Analysis of the HCT116 Genetic Interaction Matrix. This package contains the data and source code for the paper Laufer, Fischer, Billmann, Huber, Boutros, HD2013SGI, 2013.

Details

Package: HD2013SGI
Type: Package
Version: 0.0.3
License: Artistic-2.0
LazyLoad: true
Imports:
Depends: rhdf5, RColorBrewer, gplots, geneplotter, MASS, grid, hwriter, splots, igraph, abind, limma, vcd, LSD
SystemRequirements: GNU make
biocViews: Infrastructure
Built: R 2.15.1; ; 2013-02-14 12:08:09 UTC; unix

The interaction matrix can be loaded by data(Interactions, package="HD2013SGI"). Type ?Interactions to see a documentation of the interaction data.

The vignette of the package can be seen by typing >library("HD2013SGI") >vignette("HD2013SGI") It contains the complete documentation and R-code for the analysis of the data published in the original publication.

All intermediate results are precomputed and can be loaded. the following datasets are available:

featuresPerWell The screen data in screen order
datamatrixfull The phenotype data of all pairwise genetic perturbation experiments before quality control and feature selection
QueryAnnotation Annotation of all the query genes in the screen
TargetAnnotation Annotation of all target genes in the screen
stabilitySelection Results from the feature selection step
datamatrix The phenotype data of all pairwise genetic perturbation experiments after quality control and feature selection
mainEffects
estimated main effects (single knock down effects) nrOfInteractionsPerTarget number of interactions per target gene
Interactions The genetic interaction data (pi-scores, p-values, and annotation)

A number of helper functions are defined in the package and used in the vignette.

HD2013SGIorderDim hclust on one out of three dimensions of an interaction matrix
HD2013SGIHeatmapHuman plotting a heatmap of a three dimensional array of pi-scores (target genes x query genes x features)
HD2013SGImaineffects estimation main effects
HD2013SGIselectByStability feature selection to select features most stable between replicated experiments

Author(s)

Bernd Fischer

Maintainer: Bernd Fischer <[email protected]>

References

Laufer, Fischer, Billmann, Huber, Boutros, HD2013SGI, 2013.

Examples

data(Interactions, package="HD2013SGI")

Phenotypic data after quality control and feature selection

Description

Phenotypic features of pairwise genetic perturbation experiments after selection of non-redundant features and quality control. D is the 6-dimensional array of experimental measurements. Its dimensions are target genes x target siRNA designs x query genes x query siRNA designs x features x replicates. The array has a dimnames attribute, but there exists a more comprehensive annotation of target genes, query genes, and phenotypes in Anno.

Usage

data(datamatrix)

Format

The format is: List of 2 $ D : num [1:289, 1:2, 1:20, 1:2, 1:11, 1:2] -0.05334 -0.20294 -0.10123 0.33203 0.00638 ... ..- attr(*, "dimnames")=List of 6 .. ..$ targetGene : chr [1:289] "TDRD6" "PRDM11" "KDM1B" "INTS12" ... .. ..$ targetDesign: chr [1:2] "1" "2" .. ..$ queryGene : chr [1:20] "DPF2" "SMARCA1" "SMARCC1" "SMARCD2" ... .. ..$ queryDesign : chr [1:2] "1" "2" .. ..$ features : chr [1:11] "count" "cell.act.m.majoraxis" "nuc.nuc.b.q001" "nuc.0.m.eccentricity" ... .. ..$ replicate : chr [1:2] "1" "2" $ Anno:List of 3 ..$ target :'data.frame': 289 obs. of 4 variables: .. ..$ ID : chr [1:289] "B1" "B2" "B3" "B4" ... .. ..$ Symbol: chr [1:289] "TDRD6" "PRDM11" "KDM1B" "INTS12" ... .. ..$ Well : chr [1:289] "B1" "B2" "B3" "B4" ... .. ..$ group : chr [1:289] "sample" "sample" "sample" "sample" ... ..$ query :'data.frame': 20 obs. of 2 variables: .. ..$ ID : chr [1:20] "01" "02" "03" "04" ... .. ..$ Symbol: chr [1:20] "DPF2" "SMARCA1" "SMARCC1" "SMARCD2" ... ..$ phenotype: chr [1:11] "count" "cell.act.m.majoraxis" "nuc.nuc.b.q001" "nuc.0.m.eccentricity" ...

References

Laufer, Fischer et al., 2013

See Also

HD2013SGI

Examples

data(datamatrix, package="HD2013SGI")
plot(datamatrix$D[,1,1,1,1,1])

Phenotypic data before quality control and feature selection

Description

Phenotypic features of pairwise genetic perturbation experiments before selection of non-redundant features and quality control. D is the 6-dimensional array of experimental measurements. Its dimensions are target genes x target siRNA designs x query genes x query siRNA designs x features x replicates. The array has a dimnames attribute.

Usage

data(datamatrixfull)

Format

The format is: List of 1 $ D: num [1:345, 1:2, 1:20, 1:2, 1:353, 1:2] 2686 2573 2650 3000 2733 ... ..- attr(*, "dimnames")=List of 6 .. ..$ targetGene : chr [1:345] "B1" "B2" "B3" "B4" ... .. ..$ targetDesign: chr [1:2] "1" "2" .. ..$ queryGene : chr [1:20] "01" "02" "03" "04" ... .. ..$ queryDesign : chr [1:2] "1" "2" .. ..$ features : chr [1:353] "count" "nuc.0.m.cx" "nuc.0.m.cy" "nuc.0.m.majoraxis" ... .. ..$ replicate : chr [1:2] "1" "2"

References

Laufer, Fischer et al., 2013

See Also

HD2013SGI

Examples

data(datamatrixfull, package="HD2013SGI")
plot(datamatrixfull$D[,1,1,1,1,1])

Original phenotypic measurements in screen order

Description

Original phenotypic features in screen order. Anno contains the annotation for each experiment including the plate name, row, col, and field. data is a data.frame with a column for each phenotypic feature and rows as much as there are experiments in the screen.

Usage

data(featuresPerWell)

Format

The format is: List of 2 $ Anno:'data.frame': 231840 obs. of 4 variables: ..$ plate: chr [1:231840] "001CIQ01IRI" "001CIQ01IRI" "001CIQ01IRI" "001CIQ01IRI" ... ..$ row : chr [1:231840] "B" "B" "B" "B" ... ..$ col : chr [1:231840] "1" "1" "1" "1" ... ..$ field: chr [1:231840] "1" "2" "3" "4" ... $ data: num [1:231840, 1:353] 2780 3120 2242 2603 2170 ... ..- attr(*, "dimnames")=List of 2 .. ..$ : NULL .. ..$ : chr [1:353] "count" "nuc.0.m.cx" "nuc.0.m.cy" "nuc.0.m.majoraxis" ...

References

Laufer, Fischer et al., 2013

See Also

HD2013SGI

Examples

data(featuresPerWell, package="HD2013SGI")
plot(log2(featuresPerWell$data[,1]),pch=".")

Plotting heatmaps of genetic interaction scores

Description

Plotting heatmaps of three-dimensional arrays of interaction scores. Two dimensions of the array will be flattened.

Usage

HD2013SGIHeatmapHuman(x, cuts, col, colnames = TRUE, rownames = FALSE, mrow = 10, mcol = 10, cexrow = 1, cexcol = 1, border = 0.1, space = 0.05)

Arguments

x

A three dimensional array to be plotted as a heatmap.
cuts

cuts on the values of x for color coding. length(cuts) has be one larger than length(col).
col

Values of x are mapped on color definitions as defined in col using the cuts argument.
colnames

Gene names for columns of the matrix.
rownames

Gene names for rows of the matrix.
mrow

row margin for printing gene names..
mcol

column margin for printing gene names.
cexrow

cex for the row names.
cexcol

cex for the column names.
border

line width of the border.
space

spacing between elements of the third array dimension of x after flattening.

Value

Nothing is returned.

Author(s)

Bernd Fischer

References

Laufer, Fischer et al., 2013

See Also

HD2013SGI

Estimation of main effects

Description

A function to estimate main effects (single knock-down effects) in genetic interaction screens.

Usage

HD2013SGImaineffects(x, TP, TargetNeg, QueryNeg, eps = 1e-04, maxiter = 100, na.rm = TRUE)

Arguments

x

Two dimensional array.
TP

Assignment of target genes to target plates. Used to compute target main effects for each target plate separately.
TargetNeg

Negative controls within the set of target genes.
QueryNeg

Negative controls within the set of query genes.
eps

real number greater than 0. A tolerance for convergence.
maxiter

the maximum number of iterations
na.rm

logical. Should missing values be removed?

Value

neg

Effect of the negative control.
targetMainEffect

target main effects
queryMainEffect

query main effects
pi

Pairwise interaction scores (pi-scores)

Author(s)

changes applied by Bernd Fischer to the implementation of R stats functionmedpolish

References

Laufer, Fischer et al., 2013

See Also

HD2013SGI

hclust on one out of three dimensions of an interaction matrix

Description

hclust on one out of three dimensions of a three-dimensional array of interaction scores (target genes x query genes x features)

Usage

HD2013SGIorderDim(x, i)

Arguments

x

A three dimensional array to be clustered.
i

The dimension of the array along which the data is clustered.

Value

Returns a cluster hierarchy of class hclst.

Author(s)

Bernd Fischer

References

Laufer, Fischer et al., 2013

See Also

HD2013SGI

Feature selection method

Description

A function to select features that are most stable across replicated experiments

Usage

HD2013SGIselectByStability(subsample, preselect = "count", Rdim = 40, verbose = TRUE)

Arguments

subsample

This is the input data. Usually a subsample of the complete screen is enough to select the non-redundant features. subsample is a list with three elements: D is a three-dimensional array with dimensions samples x features x replicates. As samples usually 1000 to 5000 experiments are randomly selected. The function needs two replicates.

preselect

Names of the features that should be preselected, e.g. count is preselected in this screen, because of its biological interpretability and comparability to other viability-based genetic interaction screens.
Rdim

The maximum number of features to be selected.
verbose

If TRUE information about the progress and the quality of the selected features is printed.

Value

(selected = selected, correlation = correlation, ratioPositive = ratioPositive, correlationAll = correlationAll)

selected

The names of the selected features in the order as selected.
correlation

The correlation of the residual features after fitting a linear function on the previously selected features. Correlations are in same order as selected.
ratioPositive

The fraction of positively correlated features among all candidate features in each step. In same order as selected.
correlationAll

The correlation of the residual features of all candidate features in each step of the selection process.

Author(s)

Bernd Fischer

References

Laufer, Fischer et al., 2013

See Also

HD2013SGI

The genetic interaction data.

Description

The genetic interaction data. Pairwise interaction scores (piscore) are presented in a 6-dimensional array with dimensions target genes x target siRNA designs x query genes x query siRNA designs x features x replicates. BH-corrected p-values (padj) are presented in a 5-dimensional array with dimensions target genes x target siRNA designs x query genes x query siRNA designs x features. An annotation of target and query genes and of phenotypes can be found in (Anno). scale is the standard deviation measure used for normalization. At first standard deviations were computed between replicates for each experiment and afterwards the median of standard deviations was computed.

Usage

data(Interactions)

Format

The format is: List of 4 $ piscore: num [1:282, 1:2, 1:20, 1:2, 1:11, 1:2] -1.814 -2.457 -3.094 -1.448 -0.142 ... ..- attr(*, "dimnames")=List of 6 .. ..$ targetGene : chr [1:282] "TDRD6" "PRDM11" "KDM1B" "INTS12" ... .. ..$ targetDesign: chr [1:2] "1" "2" .. ..$ queryGene : chr [1:20] "DPF2" "SMARCA1" "SMARCC1" "SMARCD2" ... .. ..$ queryDesign : chr [1:2] "1" "2" .. ..$ features : chr [1:11] "count" "cell.act.m.majoraxis" "nuc.nuc.b.q001" "nuc.0.m.eccentricity" ... .. ..$ replicate : chr [1:2] "1" "2" $ scale : num [1:11] 0.214 0.125 0.139 0.193 0.207 ... $ padj : num [1:282, 1:2, 1:20, 1:2, 1:11] 0.6838 0.4167 0.0949 0.5786 0.7933 ... ..- attr(*, "dimnames")=List of 5 .. ..$ targetGene : chr [1:282] "TDRD6" "PRDM11" "KDM1B" "INTS12" ... .. ..$ targetDesign: chr [1:2] "1" "2" .. ..$ queryGene : chr [1:20] "DPF2" "SMARCA1" "SMARCC1" "SMARCD2" ... .. ..$ queryDesign : chr [1:2] "1" "2" .. ..$ features : chr [1:11] "count" "cell.act.m.majoraxis" "nuc.nuc.b.q001" "nuc.0.m.eccentricity" ... $ Anno :List of 3 ..$ target :'data.frame': 282 obs. of 4 variables: .. ..$ ID : chr [1:282] "B1" "B2" "B3" "B4" ... .. ..$ Symbol: chr [1:282] "TDRD6" "PRDM11" "KDM1B" "INTS12" ... .. ..$ Well : chr [1:282] "B1" "B2" "B3" "B4" ... .. ..$ group : chr [1:282] "sample" "sample" "sample" "sample" ... ..$ query :'data.frame': 20 obs. of 2 variables: .. ..$ ID : chr [1:20] "01" "02" "03" "04" ... .. ..$ Symbol: chr [1:20] "DPF2" "SMARCA1" "SMARCC1" "SMARCD2" ... ..$ phenotype: chr [1:11] "count" "cell.act.m.majoraxis" "nuc.nuc.b.q001" "nuc.0.m.eccentricity" ...

References

Laufer, Fischer et al., 2013

See Also

HD2013SGI

Examples

data(Interactions, package="HD2013SGI")
plot(Interactions$piscore[,1,"SUV39H1",1,"cell.act.m.majoraxis",1])
print(names(which(Interactions$padj[,1,"SUV39H1",1,
                                    "cell.act.m.majoraxis"] <= 0.01)))

Estimated main effects

Description

Estimated main effects (single knock-down effects) for target and query genes. Additional overall effects for each phenotype are contained. The dataset contains an annotation of target genes, query genes, and phenotypes.

Usage

data(mainEffects)

Format

The format is: List of 4 $ target: num [1:289, 1:2, 1:2, 1:11, 1:2] -0.31065 -0.32253 -0.08466 -0.00367 -0.60867 ... ..- attr(*, "dimnames")=List of 5 .. ..$ targetGene : chr [1:289] "TDRD6" "PRDM11" "KDM1B" "INTS12" ... .. ..$ targetDesign: chr [1:2] "1" "2" .. ..$ queryDesign : chr [1:2] "1" "2" .. ..$ features : chr [1:11] "count" "cell.act.m.majoraxis" "nuc.nuc.b.q001" "nuc.0.m.eccentricity" ... .. ..$ replicate : chr [1:2] "1" "2" $ query : num [1:2, 1:20, 1:2, 1:11, 1:2] 0.277 0.265 0.235 0.226 1.165 ... ..- attr(*, "dimnames")=List of 5 .. ..$ targetDesign: chr [1:2] "1" "2" .. ..$ queryGene : chr [1:20] "DPF2" "SMARCA1" "SMARCC1" "SMARCD2" ... .. ..$ queryDesign : chr [1:2] "1" "2" .. ..$ features : chr [1:11] "count" "cell.act.m.majoraxis" "nuc.nuc.b.q001" "nuc.0.m.eccentricity" ... .. ..$ replicate : chr [1:2] "1" "2" $ overall : num [1:2, 1:2, 1:11, 1:2] 0.3685 0.4638 0.3331 0.487 -0.0985 ... ..- attr(*, "dimnames")=List of 4 .. ..$ targetDesign: chr [1:2] "1" "2" .. ..$ queryDesign : chr [1:2] "1" "2" .. ..$ features : chr [1:11] "count" "cell.act.m.majoraxis" "nuc.nuc.b.q001" "nuc.0.m.eccentricity" ... .. ..$ replicate : chr [1:2] "1" "2" $ Anno :List of 3 ..$ target :'data.frame': 289 obs. of 4 variables: .. ..$ ID : chr [1:289] "B1" "B2" "B3" "B4" ... .. ..$ Symbol: chr [1:289] "TDRD6" "PRDM11" "KDM1B" "INTS12" ... .. ..$ Well : chr [1:289] "B1" "B2" "B3" "B4" ... .. ..$ group : chr [1:289] "sample" "sample" "sample" "sample" ... ..$ query :'data.frame': 20 obs. of 2 variables: .. ..$ ID : chr [1:20] "01" "02" "03" "04" ... .. ..$ Symbol: chr [1:20] "DPF2" "SMARCA1" "SMARCC1" "SMARCD2" ... ..$ phenotype: chr [1:11] "count" "cell.act.m.majoraxis" "nuc.nuc.b.q001" "nuc.0.m.eccentricity" ...

References

Laufer, Fischer et al., 2013

See Also

HD2013SGI

Examples

data(mainEffects, package="HD2013SGI")
print(dim(mainEffects$target))
plot(mainEffects$target[,1,1,1,1])

Number of interactions per target gene

Description

Number of genetic interactions per target gene.

Usage

data(nrOfInteractionsPerTarget)

Format

The format is: int [1:282] 1 3 0 2 0 3 1 1 0 1 ...

References

Laufer, Fischer et al., 2013

See Also

HD2013SGI

Examples

data(nrOfInteractionsPerTarget, package="HD2013SGI")
plot(nrOfInteractionsPerTarget)

Annotation of all query genes in the screen

Description

Annotation of all query genes in the screen.

Usage

data(QueryAnnotation)

Format

A data frame with 20 observations on the following 2 variables.

ID

a character vector

Symbol

a character vector

References

Laufer, Fischer et al., 2013

See Also

HD2013SGI

Examples

data(QueryAnnotation, package="HD2013SGI")
print(QueryAnnotation$Symbol)

Results from the feature selection method

Description

Results from the feature selection method.

Usage

data(stabilitySelection)

Format

The format is: List of 4 $ selected : chr [1:25] "count" "cell.act.m.majoraxis" "nuc.nuc.b.q001" "nuc.0.m.eccentricity" ... $ correlation : num [1:25] 0.917 0.972 0.938 0.928 0.896 ... $ ratioPositive : num [1:25] 1 1 0.947 0.942 0.937 ... $ correlationAll: List of 25 ..$ : Named num [1:227] 0.917 0.884 0.93 0.897 0.882 ... .. ..- attr(*, "names")= chr [1:227] "count" "nuc.0.m.majoraxis" "nuc.0.m.eccentricity" "nuc.0.s.area" ... ..$ : Named num [1:226] 0.884 0.931 0.899 0.887 0.9 ... .. ..- attr(*, "names")= chr [1:226] "nuc.0.m.majoraxis" "nuc.0.m.eccentricity" "nuc.0.s.area" "nuc.0.s.perimeter" ... ..$ : Named num [1:225] 0.884 0.934 0.884 0.882 0.883 ... .. ..- attr(*, "names")= chr [1:225] "nuc.0.m.majoraxis" "nuc.0.m.eccentricity" "nuc.0.s.area" "nuc.0.s.perimeter" ... ...

Details

selected is a vector of the selected feature namess. correlation are the Pearson correlation coefficients of the residual features. ratioPositive is the fraction of positively correlated features among all candidate features for selection. correlationAll contains a vector of correlations of the residual features of all candidate features for each step in the selection process.

References

Laufer, Fischer et al., 2013

See Also

HD2013SGI

Examples

data(stabilitySelection, package="HD2013SGI")
barplot(stabilitySelection$correlation,
        names.arg=stabilitySelection$selected,las=2)
barplot(stabilitySelection$ratioPositive-0.5, offset=0.5,
        names.arg=stabilitySelection$selected,las=2)

Annotation of all target genes in the screen

Description

Annotation of the target genes on one target plate. It includes an ENSEMBL gene identifier, the HUGO name, the position on the plate (well), and the group of the target siRNA (sample or control).

Usage

data(TargetAnnotation)

Format

A data frame with 345 observations on the following 4 variables.

ID

a character vector

Symbol

a character vector

Well

a character vector

group

a character vector

References

Laufer, Fischer et al., 2013

See Also

HD2013SGI

Examples

data(TargetAnnotation, package="HD2013SGI")
print(TargetAnnotation$Symbol)