Package {MultiSEp}

The output of mts_formatMatrix() function for an example dataset.

Description

List of 23 dataframes where each element corresponds to a gene. The mutation impact data per sample is given for each gene, correctly formatted for analysis with the mts_patternDetection function.

Usage

data("Formatted_MUTimpactMatrix")

Format

A data frame with 3 columns: 'Samples', 'Impact' and 'impact Group'.

Examples

data(Formatted_MUTimpactMatrix)

Example mutation impact matrix

Description

A binary/numeric matrix in gene by sample format, where row names correspond to genes (or probesets) and columns are sample names. The matrix cells must have values of '1' or '2' to represent mutation impact (1 for HIGH impact, 2 for LOW impact). This matrix has been adapted from DepMap (Tsherniak et al. 2017), the numeric values in the matrix were generated pseudo-randomly.

Usage

data("MUT_impactMatrix")

Format

A data frame with 23 observations (genes) on 26 variables (cell lines).

Source

DepMap 2024Q2 doi:10.25452/figshare.plus.25880521.v1

References

Tsherniak A., Vazquez F., Montgomery P.G., Weir B.A., Kryukov G., Cowley G.S., Gill S., Harrington W.F., Pantel S., Krill-Burger J.M. et al. . Defining a cancer dependency map. Cell. 2017; 170:564-576. doi:10.1016/j.cell.2017.06.010

Examples

data(MUT_impactMatrix)
MUT_impactMatrix[1:10,1:10]

Example CRISPR score input matrix

Description

Log2 CRISPR scores where row names correspond to genes and column names correspond to samples.

Usage

data("depMapCRISPRscores_subset")

Format

A data frame with 50 observations (genes) across 1054 variables (cell lines).

Source

DepMap 2024Q2 doi:10.25452/figshare.plus.25880521.v1

References

Tsherniak A., Vazquez F., Montgomery P.G., Weir B.A., Kryukov G., Cowley G.S., Gill S., Harrington W.F., Pantel S., Krill-Burger J.M. et al. . Defining a cancer dependency map. Cell. 2017; 170:564-576. doi:10.1016/j.cell.2017.06.010

Examples

data(depMapCRISPRscores_subset)
depMapCRISPRscores_subset[1:10,1:10]

Example mutation data input matrix

Description

Discrete mutation values where row names correspond to genes and column names correspond to samples. Wild type denoted by 'WT' and mutations are indicated by amino acid changes.

Usage

data("depMapMUT_small")

Format

A data frame with 6 observations (genes) across 1299 variables (cell lines).

Source

DepMap 2024Q2 doi:10.25452/figshare.plus.25880521.v1

References

Tsherniak A., Vazquez F., Montgomery P.G., Weir B.A., Kryukov G., Cowley G.S., Gill S., Harrington W.F., Pantel S., Krill-Burger J.M. et al. . Defining a cancer dependency map. Cell. 2017; 170:564-576. doi:10.1016/j.cell.2017.06.010

Examples

data(depMapMUT_small)
depMapMUT_small[1:10,1:10]

Example mutation data input matrix

Description

Discrete mutation values where row names correspond to genes and column names correspond to samples. Wild type denoted by 'WT' and mutations are indicated by amino acid changes.

Usage

data("depMapMUT_subset")

Format

A data frame with 132 observations (genes) across 1738 variables (cell lines).

Source

DepMap 2024Q2 doi:10.25452/figshare.plus.25880521.v1

References

Tsherniak A., Vazquez F., Montgomery P.G., Weir B.A., Kryukov G., Cowley G.S., Gill S., Harrington W.F., Pantel S., Krill-Burger J.M. et al. . Defining a cancer dependency map. Cell. 2017; 170:564-576. doi:10.1016/j.cell.2017.06.010

Examples

data(depMapMUT_subset)
depMapMUT_subset[1:10,1:10]

Example tissue data input matrix

Description

A data frame with 789 observations (samples) across 2 variables (tissues).

Usage

data("depMapTissue_subset")

Format

A data frame where column 1 corresponds to the sample ID and column 2 corresponds to the tissue value.

cell_line

a character vector

tissue

a character vector

Source

DepMap 2024Q2 doi:10.25452/figshare.plus.25880521.v1

References

Tsherniak A., Vazquez F., Montgomery P.G., Weir B.A., Kryukov G., Cowley G.S., Gill S., Harrington W.F., Pantel S., Krill-Burger J.M. et al. . Defining a cancer dependency map. Cell. 2017; 170:564-576. doi:10.1016/j.cell.2017.06.010

Examples

data(depMapTissue_subset)
depMapTissue_subset[1:10,]

Example gene expression input matrix

Description

Log2 gene expression data for 11 observations (genes) across 1299 variables (cell lines).

Usage

data("depMapXPR_small")

Format

A data frame where rownames correspond to genes and column names correspond to samples.

Source

DepMap 2024Q2 doi:10.25452/figshare.plus.25880521.v1

References

Tsherniak A., Vazquez F., Montgomery P.G., Weir B.A., Kryukov G., Cowley G.S., Gill S., Harrington W.F., Pantel S., Krill-Burger J.M. et al. . Defining a cancer dependency map. Cell. 2017; 170:564-576. doi:10.1016/j.cell.2017.06.010

Examples

data(depMapXPR_small)
depMapXPR_small[1:10,1:10]

Example gene expression input matrix

Description

Log2 gene expression data for 50 observations (genes) across 1304 variables (cell lines).

Usage

data("depMapXPR_subset")

Format

A data frame where rownames correspond to genes and column names correspond to samples.

Source

DepMap 2024Q2 doi:10.25452/figshare.plus.25880521.v1

References

Tsherniak A., Vazquez F., Montgomery P.G., Weir B.A., Kryukov G., Cowley G.S., Gill S., Harrington W.F., Pantel S., Krill-Burger J.M. et al. . Defining a cancer dependency map. Cell. 2017; 170:564-576. doi:10.1016/j.cell.2017.06.010

Examples

data(depMapXPR_subset)
depMapXPR_subset[1:10,1:10]

Example results for the mts_Mutation function.

Description

A data frame containing example predicted gene dependencies produced by the mts_Mutation function (which takes as input gene expression and mutation data).

Usage

data("mRes")

Format

A data frame with 3039 observations on the following 5 variables.

mRNA_gene

a character vector

mutation_gene

a character vector

tissue

a character vector

chiSqPvalue

a numeric vector

mutation_per_mode

a character vector

References

Wappett et al. (2021) [Nucleic Acids Research 49, W613-W618] doi:10.1093/nar/gkab338

Examples

data(mRes)
mRes[1:10,]

Clusters of samples (cell lines) from the mts_mixModelCluster function for a small dataset of DepMap CRISPR scores.

Description

List of 5 data frames where each element corresponds to a gene.

Usage

data("mixModelClusters_depMapCRISPR")

Format

Each data frame in the list has 3 columns: gene ID, log2 gene expression and cluster assignment. Also provides an example of a gene, EIF1AX, that was not able to receive cluster assignments.

Source

Derived from analysis of data taken from DepMap 2024Q2 doi:10.25452/figshare.plus.25880521.v1

References

Tsherniak A., Vazquez F., Montgomery P.G., Weir B.A., Kryukov G., Cowley G.S., Gill S., Harrington W.F., Pantel S., Krill-Burger J.M. et al. . Defining a cancer dependency map. Cell. 2017; 170:564-576. doi:10.1016/j.cell.2017.06.010

Examples

data(mixModelClusters_depMapCRISPR)

Clusters output from the mts_mixModelCluster function for a toy DepMap expression dataset.

Description

List of 5 data frames where each element corresponds to a gene.

Usage

data("mixModelClusters_depMapXPR")

Format

A data frame with 3 columns: gene ID, log2 gene expression and cluster assignment.

Source

Derived from analysis of data taken from DepMap 2024Q2 doi:10.25452/figshare.plus.25880521.v1

References

Tsherniak A., Vazquez F., Montgomery P.G., Weir B.A., Kryukov G., Cowley G.S., Gill S., Harrington W.F., Pantel S., Krill-Burger J.M. et al. . Defining a cancer dependency map. Cell. 2017; 170:564-576. doi:10.1016/j.cell.2017.06.010

Examples

data(mixModelClusters_depMapXPR)

Clusters output from the mts_mixModelCluster function for a DepMap expression data that is required for the vignette.

Description

List of 9 data frames where each element corresponds to a gene.

Usage

data("mixModelClusters_vignette")

Format

Each data frame has 3 columns: gene ID, log2 gene expression and cluster assignment.

Source

Derived from analysis of data taken from DepMap 2024Q2 doi:10.25452/figshare.plus.25880521.v1

References

Tsherniak A., Vazquez F., Montgomery P.G., Weir B.A., Kryukov G., Cowley G.S., Gill S., Harrington W.F., Pantel S., Krill-Burger J.M. et al. . Defining a cancer dependency map. Cell. 2017; 170:564-576. doi:10.1016/j.cell.2017.06.010

Examples

data(mixModelClusters_vignette)

Investigate Gene Dependency Relationships with CRISPR and gene expression data

Description

Performs statistical evaluation of gene dependency relationships for CRISPR scores between consecutive mRNA expression clusters.

Usage

mts_Crispr(resultList = resultList, exprsMatrix = exprsMatrix,
crisprMatrix = crisprMatrix, fcVal = fcVal, pVal = pVal, cores = cores)

Arguments

Details

Gene expression clusters enable partitioning of effect scores from CRISPR screen data, and therefore may reveal pairwise gene dependencies. Statistical evaluation of dependency relationships is carried out by calculating a log2 fold-change and two-tailed T-test p-value between the CRISPR scores for each consecutive pair of mRNA expression clusters. The user can define the fold change (default:-0.1) and p-value (default: 0.1) cutoff values. Q-values are also calculated; we strongly recommend taking the q-values as a baseline for determining statistical significance, indeed it is crucially important to correct for multiple hypothesis testing to ensure the statistical validity of results.

Value

A data frame containing the significantly enriched gene expression and CRISPR gene pair results. The data frame contains the following columns:

See Also

mts_clusterAvg

Examples

data(depMapXPR_subset)
data(depMapCRISPRscores_subset)
data(multisepList)

resultListSmall <- lapply(multisepList["DNAJC15"], function(g) {
    keep <- intersect(rownames(g[[1]]), rownames(depMapCRISPRscores_subset))
    g[[1]] <- g[[1]][keep, , drop = FALSE]
    g
  })

mCrisprOut <- mts_Crispr(
  resultList = resultListSmall,
  exprsMatrix = depMapXPR_subset, 
  crisprMatrix = depMapCRISPRscores_subset
)

Predict tissue-specific gene dependency relationships, evaluating differences in CRISPR scores for gene A between expression clusters for gene B (SynLeGG).

Description

Subsets the output from the mts_Crispr function according to tissue type and performs the analysis again in a tissue specific manner. Running with negative fold-change values (<0) allows prediction of synthetic lethal relationships, specifying positive fold-change values (>0) corresponds to 'induced dependency' relationships.

Usage

mts_CrisprTS(resultList = resultList, crisprMatrix = crisprMatrix, fcVal = fcVal, 
pVal = pVal, cores = cores, tissueMatrix = tissueMatrix, allDisRes = allDisRes)

Arguments

Value

A data frame containing the significantly enriched gene expression and CRISPR gene pair results. The data frame contains the following columns:

References

Wappett et al. (2021) [Nucleic Acids Research 49, W613-W618] doi:10.1093/nar/gkab338

Examples

data(depMapCRISPRscores_subset)
data(depMapTissue_subset)
data(multisepList)
data(multisepCrisprList)
allDis <- multisepCrisprList[multisepCrisprList$crispr_gene %in%
                               rownames(depMapCRISPRscores_subset), ][1:15, ]

mCrisprTS <- mts_CrisprTS(
  resultList = multisepList,
  crisprMatrix = depMapCRISPRscores_subset, 
  tissueMatrix = depMapTissue_subset,
  fcVal = -0.1, 
  pVal = 0.25, 
  allDisRes = allDis)

CRISPR and gene expression gene-dependency prediction (SynLeGG)

Description

Run the CRISPR and expression gene-dependency pipeline, optionally with tissue data. Minimally, an expression matrix and a CRISPR matrix must be provided.

Usage

mts_GeneDepCrispr(exprsMatrix = exprsMatrix, crisprMatrix = crisprMatrix,
cores = cores, fcVal = fcVal, pVal = pVal, tissueMatrix = tissueMatrix)

Arguments

Value

A list containing up to two tables:

Examples

data(depMapCRISPRscores_subset)
data(depMapXPR_subset)
data(depMapTissue_subset)
exprsSub  <- depMapXPR_subset["PCYT1B",]
crisprSub <- depMapCRISPRscores_subset["PCYT1A",]
mts_GeneDepCrispr(
exprsMatrix = exprsSub, crisprMatrix = crisprSub,
cores = 1, fcVal = -0.5, pVal = 0.25, tissueMatrix = depMapTissue_subset)

CRISPR vs gene expression induced dependency pipeline (SynLeGG)

Description

Run the MultiSEp CRISPR vs gene expression induced dependency pipeline, optionally with tissue data. Minimally, a gene expression matrix and a CRISPR matrix must be provided.

Usage

mts_GeneDepID(exprsMatrix = exprsMatrix, crisprMatrix = crisprMatrix,
cores = cores, fcVal = fcVal, pVal = pVal, tissueMatrix = tissueMatrix)

Arguments

Value

A list containing up to 2 tables:

Examples

data(depMapCRISPRscores_subset)
data(depMapXPR_subset)
data(depMapTissue_subset)
mtsGeneDep_ID_Results <- mts_GeneDepID(exprsMatrix=depMapXPR_subset[1:3,], 
crisprMatrix=depMapCRISPRscores_subset[1:10,], cores=1, fcVal=0.1, pVal=0.25,
tissueMatrix=depMapTissue_subset)

Mutation vs gene expression gene-dependency prediction

Description

Run the MultiSEp mutation gene-dependency pipeline, with tissue data. As a minimum an expression matrix, mutation matrix and a tissue matrix must be provided (please see details below).

Usage

mts_GeneDepMutation(exprsMatrix = exprsMatrix, mutMatrix= mutMatrix,
cores = cores, pVal = pVal, tissueMatrix = tissueMatrix)

Arguments

Value

A data frame containing the results of MultiSEp analysis of mutations by gene expression clusters, with the following columns:

Examples

data(depMapXPR_subset)
data(depMapMUT_subset)
data(depMapTissue_subset)

mtsGeneDepMutResults <- mts_GeneDepMutation(
  exprsMatrix=depMapXPR_subset[1,],
  tissueMatrix = depMapTissue_subset,
  mutMatrix = depMapMUT_subset[1:2, ], pVal=0.1)

Predict 'induced dependency' relationships, evaluating differences in CRISPR scores for gene A between expression clusters for gene B

Description

Gene expression clusters enable partitioning of effect scores from CRISPR screen data, and therefore may reveal pairwise gene dependencies. mts_InducedDependency performs statistical evaluation of induced dependency relationships (enrichment) by calculating a log2 fold-change and two-tailed T-test p-value between the CRISPR scores for each consecutive pair of mRNA expression clusters. 'Induced dependency' relationships in this function are defined by lower CRISPR scores (corresponding to reduced cell survival/growth) with higher gene expression values; corresponding to filtering by positive fold-change values (>0). Q-values are also calculated (using false discovery rate correction). We strongly recommend taking the q-values as a baseline for determining statistical significance, indeed it is crucially important to correct for multiple hypothesis testing to ensure the statistical validity of results.

Usage

mts_InducedDependency(resultList = resultList, exprsMatrix = exprsMatrix,
crisprMatrix = crisprMatrix, fcVal = fcVal, pVal = pVal, cores = cores)

Arguments

Details

Similar to the methodology applied for CRISPR analysis in the www.overton-lab.uk/synlegg resource (Wappett et al. 2021); developments include an improved approach for discovery of the gene expression clusters.

Value

A data frame containing results for gene expression and CRISPR gene pairs that pass the p-value and log2 fold-change threshold values used. The data frame contains the following columns:

References

Wappett et al. (2021) [Nucleic Acids Research 49, W613-W618] doi:10.1093/nar/gkab338

See Also

mts_clusterAvg

Examples

data(depMapXPR_subset)
data(depMapCRISPRscores_subset)
data(multisepList)

resultListSmall <- lapply(multisepList["DNAJC15"], function(g) {
    keep <- intersect(rownames(g[[1]]), rownames(depMapCRISPRscores_subset))
    g[[1]] <- g[[1]][keep, , drop = FALSE]
    g
  })

mIDOut <-  mts_InducedDependency(
    resultList   = resultListSmall,
    exprsMatrix  = depMapXPR_subset,
    crisprMatrix = depMapCRISPRscores_subset
  )

Integrate gene cluster assignments with mutation data

Description

MultiSEp detects mutations that are enriched in gene clusters (for example, from mRNA expression); providing insight into the genetic background relevant to candidate 'Achilles Heel' relationships suggested by analysis of the CRISPR screen and transcriptome data. A chi-squared test is performed for each gene to derive p-values for the enrichment of mutation classes across the gene expression clusters on a per-tissue basis.

Usage

mts_Mutation(resultList = resultList, mutMatrix = mutMatrix, tissueMatrix = tissueMatrix, 
cores = cores, pVal = pVal)

Arguments

Value

A data frame containing the results of MultiSEp analysis of mutations by gene expression clusters, with the following columns:

See Also

mts_mixModelCluster

Examples

data(depMapXPR_subset)
data(depMapMUT_subset)
data(depMapTissue_subset)
mixModelClusters <- mts_mixModelCluster(dataMatrix = depMapXPR_subset[1,])

multisepMut <- mts_Mutation(
  resultList = mixModelClusters[1],
  mutMatrix = depMapMUT_subset, 
  tissueMatrix = depMapTissue_subset,
  pVal = 0.01
)

Determine average CRISPR score values for gene expression clusters

Description

MultiSEp applies regularised Gaussian Mixture Modelling (GMM) with expectation-maximisation to discover clusters, or modes, from functional genomics data; for example gene expression (such as RNA-seq), or CRISPR scores. Cardinality (the number of clusters) is determined by regularisation with Bayesian Information Criterion, considering at least two and up to five clusters. Each sample is assigned a probability of belonging to each cluster by MultiSEp, for example some samples may be positioned in a region of overlap between two clusters formed from the mixture of Gaussians used to model the distribution of the functional genomics data. The mts_clusterAvg function generates Gaussian mixture models for gene expression data, assigning samples to clusters for each gene. Subsequently, the average CRISPR score value is calculated per cluster of samples for all of the genes with CRISPR data. Accordingly, each gene with expression data is paired with all of the genes with CRISPR data across the samples analysed and these results provide the basis for quantification of the differences in CRISPR scores between gene expression clusters in downstream functions.

Usage

mts_clusterAvg(exprsMatrix, crisprMatrix, cores)

Arguments

Value

A list of sublists corresponding to each gene in the exprsMatrix object. Each sublist has the following components:

References

Wappett et al. (2021) [Nucleic Acids Research 49, W613-W618] doi:10.1093/nar/gkab338

See Also

mts_plotCRISPRGeneCluster, mts_Crispr, mts_CrisprTS, mts_InducedDependency

Examples

data(depMapXPR_subset)
data(depMapCRISPRscores_subset)

mR1 <- mts_clusterAvg(
  exprsMatrix=depMapXPR_subset[1,],
  crisprMatrix=depMapCRISPRscores_subset[1:10],
  cores=1
)

Partition CRISPR gene effect scores

Description

Partitions the CRISPR gene effect scores using the supplied threshold. For example a score of -0.5 is the default threshold, which aims to partition samples into 'lethal' (representative of cell death or stasis) and 'non-lethal' groups.

Usage

mts_crisprPartition(
  dataMatrix, 
  NumSampleThreshold = 20,
  partitionThreshold=-0.5,
  verbose=FALSE,
  cores)

Arguments

Value

A list object for each gene in the dataMatrix object

See Also

mts_patternDetection

Examples

data("depMapCRISPRscores_subset")
crisprClusters_thresholded <- mts_crisprPartition(
  dataMatrix=depMapCRISPRscores_subset[1:5,], cores=1
)

Format a matrix into a list of data frames that capture gene functional status

Description

This function transforms an 'impact matrix' into a list of data frames, where each gene is represented by a separate data frame to provide information on the sample name, functional impact status (e.g. high/low impact mutations; high/low methylation levels) and impact group (1 for HIGH, 2 for LOW). This function should be used to format any gene by sample matrix to be used as input to the mts_patternDetection function in order to predict gene dependency relationship patterns.

Usage

mts_formatMatrix(matrix, cores) 

Arguments

Details

Formats an input matrix into the correct format for input into the mts_patternDetection() function.

Value

A list of dataframes, where each gene corresponds to a separate dataframe and contains the impact groups for samples (cell lines). Each dataframe has the following three columns:

See Also

mts_patternDetection

Examples

data("MUT_impactMatrix")
Formatted_MUTimpactMatrix = mts_formatMatrix(matrix = MUT_impactMatrix, cores = 1)

Generate .RData files containing subsets of gene pairs

Description

Generates .RData files ('chunks') containing subsets of gene pairs. These subsets are useful for MultiSEp applications in high-performance computing (HPC) environments, where each file ('chunk') may be analysed by one task in an array job. This function takes one or two list objects containing cluster assignments and generates .RData files containing subsets of the full gene pair list.

Each file contains a portion of the gene pairs that can be processed independently, enabling mts_patternDetection to run in parallel across array jobs. This facilitates scalable analysis of large gene pair sets.

Usage

mts_genepairsChunkGeneration(
  genepairs = NULL, 
  mixModelClusters1,
  mixModelClusters2=NULL, 
  num_tasks,
  output_dir = tempdir(), 
  cores
)

Arguments

Value

Splits the gene pairs into num_tasks chunks and writes each to the output directory as a ‘genepairs_chunk_<i>.RData’ file. Returns a list of length num_tasks whose elements are each NULL.

See Also

mts_mixModelCluster, mts_mixModelCluster_XPR, mts_crisprPartition, mts_formatMatrix, mts_patternDetection

Examples

data(depMapXPR_subset)
mixModelClusters = mts_mixModelCluster_XPR(
  dataMatrix = depMapXPR_subset[1:5,],
  cores=1
)

# Using a custom output directory to save the .RData files (chunks)
# Here we write to a temporary directory for the example
output_dir <- tempdir()

# Generate Gene Pairs and Save to 3 Files
mts_genepairsChunkGeneration(
  mixModelClusters1 = mixModelClusters,
  num_tasks = 3,
  output_dir = output_dir
)

Cluster continuous data with Gaussian Mixture Modelling

Description

Assigns samples into clusters by unsupervised analysis of continuous values for a given gene, flexibly capturing between two to five different sub-populations of cell lines. Input data could be transcriptome, proteome or other continuous data that informs on gene functional status. Each sample (for example, a cell line) is assigned a probability of belonging to each cluster.

Usage

mts_mixModelCluster(dataMatrix, cores)

Arguments

Details

Gaussian mixture modelling with Expectation-Maximisation (EM) discovers clusters, or modes, where cardinality is determined by Bayesian Information Criterion regularisation. The method from Lubbock et al. 2013 was adapted for application to gene dependency prediction. Contiguous clusters are ensured and the ordering of clusters is taken from the value one standard deviation below the mean rather than the mean value. For example, occasionally there is overlap between clusters determined by EM such that the samples contained in one cluster encompass the range of values within a second cluster; hence samples from the first cluster are both below the minimum value and above the maximum value for any sample in the second cluster. Downstream analysis based upon the clusters, for example construction of contingency tables, assumes contiguous clusters and this property is ensured by assigning cluster boundaries with the average of minimum and maximum values from neighbouring clusters.

Value

A list object for each gene (or molecular species e.g. proteoform) in the dataMatrix object

References

Lubbock et al. (2013) TMA Navigator: network inference, patient stratification and survival analysis with tissue microarray data [Nucleic Acids Research 41, W562-W568] doi:10.1093/nar/gkt529

See Also

mts_patternDetection, mts_genepairsChunkGeneration, mts_plotClusterDistribution, mts_genepairsChunkGeneration,

mts_plotMutation

Examples

data("depMapXPR_subset")
mixModelClusters <- mts_mixModelCluster(dataMatrix = depMapXPR_subset[1,])

Cluster gene expression data with Gaussian Mixture Modelling

Description

Assigns samples into clusters by unsupervised analysis of continuous values for a given gene, flexibly capturing between two to five different sub-populations of samples. This function is designed for RNA-seq data but may be adapted to other data types. Each sample (for example, a cell line) is assigned a probability of belonging to each cluster. Filtering steps include ensuring a gene is present in an adequate number of samples with an expression value above the background.

Usage

mts_mixModelCluster_XPR(
  dataMatrix,
  GeneXPRthresh = 3.321928,
  NumSampleThresh = 20,
  cores
)

Arguments

Details

Gaussian mixture modelling with Expectation-Maximisation (EM) discovers clusters, or modes, where cardinality is determined by Bayesian Information Criterion regularisation. The method from Lubbock et al. 2013 [Nucleic Acids Research 41, W562-W568] was adapted for application to gene dependency prediction. Contiguous clusters are ensured and the ordering of clusters is taken from the value one standard deviation below the mean rather than the mean value. For example, occasionally there is overlap between clusters determined by EM such that the samples contained in one cluster encompass the range of values within a second cluster; hence samples from the first cluster are both below the minimum value and above the maximum value for any sample in the second cluster. Downstream analysis based upon the clusters, for example construction of contingency tables, assumes contiguous clusters and this property is ensured by assigning cluster boundaries with the average of minimum and maximum values from neighbouring clusters.

Value

A list object for each gene (or molecular species e.g. proteoform) in the dataMatrix object

References

Lubbock et al. (2013) TMA Navigator: network inference, patient stratification and survival analysis with tissue microarray data [Nucleic Acids Research 41, W562-W568] doi:10.1093/nar/gkt529

See Also

mts_patternDetection, mts_formatMatrix

Examples

data("depMapXPR_subset")
mixModelClusters <- mts_mixModelCluster_XPR(dataMatrix = depMapXPR_subset[1,])

Predict synthetic lethality and other dependency relationships from mutually exclusive loss of function patterns

Description

Compare any 'omics datatype either against itself (e.g. XPR_vs_XPR) or against another data-type (e.g. XPR_vs_CRISPR). Analysis of dependency relationships across many different data types is possible. For example: mutation, RNA-seq, proteomics, methylation data, histone marks, CRISPR scores, drug sensitivity, potentially metabolomics etc. This function takes as input either one dataset (for symmetrical analysis such as XPR_vs_XPR) or two datasets (non-symmetrical analysis such as XPR_vs_mutation).

Usage

mts_omics(
  dataMatrix = NULL,
  dataMatrix2 = NULL,
  mixModelClusters1 = NULL,
  mixModelClusters2 = NULL,
  categorical1 = FALSE,
  categorical2 = FALSE,
  genepairs = NULL,
  SyntheticLethalityPrediction = TRUE,
  p_adjustMethod = c("BY","BH"),
  qVal,
  directionality = c("depletion", "enrichment"),
  include_reverse_pairs = TRUE,
  effectsize = TRUE,
  effectsize_threshold = NULL,
  verbose = FALSE,
  cores)

Arguments

Value

A data frame with one row per evaluated gene pair, containing the gene names (Gene1, Gene2), the observed count (Actual_Count), the number of cluster combinations, the sample count, the expected count (Expected_Count), and the p_value and false-discovery-rate q_value (plus an Effect_Size column when effectsize = TRUE). An empty data frame with these columns is returned when no gene pair is valid.

See Also

mts_mixModelCluster, mts_mixModelCluster_XPR, mts_patternDetection

Examples

data(mixModelClusters_depMapXPR)
Unidirectional_depletion_SL_XPRvsXPR = mts_omics(
  mixModelClusters1 = mixModelClusters_depMapXPR,
  qVal=1,
  cores=1
)

data(depMapXPR_subset)
data(depMapCRISPRscores_subset)
Bidirectional_enrichment_SL_XPRvsCRISPR = mts_omics(
  dataMatrix = depMapCRISPRscores_subset[3,],
  dataMatrix2 = depMapXPR_subset[2,],
  directionality = "enrichment",
  qVal=1,
  cores=1
)

GDR_patternAnalysis_unidirectional_depletion_XPRvsXPR = mts_omics(
  dataMatrix = depMapXPR_subset[1:2,],
  SyntheticLethalityPrediction = FALSE,
  effectsize=FALSE,
  qVal=1,
  cores=1
)


# Mutation (categorical) with gene expression.
# depMapMUT_small contains string variant calls ("WT" or e.g. "p.R167W");
# convert to the 1/2 impact convention expected by mts_formatMatrix().


data(depMapMUT_small)
data(depMapXPR_small)
mutImpact <- as.data.frame(
  ifelse(as.matrix(depMapMUT_small) == "WT", 2L, 1L),
  stringsAsFactors = FALSE, check.names = FALSE)
rownames(mutImpact) <- rownames(depMapMUT_small)

SL_mut_vs_xpr = mts_omics(
  dataMatrix   = mutImpact,
  dataMatrix2  = depMapXPR_small[3,],
  categorical1 = TRUE,
  directionality = "depletion",
  qVal         = 1,
  effectsize   = FALSE,
  cores        = 1
)

# Or with pre-computed mutation 'clusters'
mutClusters <- mts_formatMatrix(matrix = mutImpact, cores = 1)
SL_mut_vs_xpr_pre = mts_omics(
  mixModelClusters1 = mutClusters,
  dataMatrix2       = depMapXPR_small[3,],
  directionality    = "depletion",
  qVal              = 1,
  effectsize        = FALSE,
  cores             = 1
)


# Using only mutation (or other categorical) data
data(depMapMUT_small)
mutImpactSmall <- as.data.frame( # make a binary matrix
  ifelse(as.matrix(depMapMUT_small) == "WT", 2L, 1L),
  stringsAsFactors = FALSE, check.names = FALSE)
rownames(mutImpactSmall) <- rownames(depMapMUT_small)

mutClasses <- mts_formatMatrix(matrix = mutImpactSmall[,1:5], cores = 1)

SL_mut_only = mts_omics(
  mixModelClusters1 = mutClasses,
  directionality    = "depletion",
  qVal              = 1,
  effectsize        = FALSE,
  cores             = 1
  )

Visualise pairwise gene expression data with MultiSEp cluster boundaries.

Description

Produces a scatterplot of two selected genes on the x and y axis, the boundaries between MultiSEp clusters are shown as blue lines.

Usage

mts_pairwiseExpressionPlot(exprsMatrix, gene1, gene2)

Arguments

Value

Produces a scatter plot (base R graphics) of the two genes' log2 expression, with mixture-model cluster boundaries drawn as reference lines.

Examples

data(depMapXPR_subset)
mts_pairwiseExpressionPlot(exprsMatrix = depMapXPR_subset, gene1 = "STX2", gene2="VRK2")

Predict Gene Dependency Relationship Patterns, including Synthetic Lethality

Description

Enumerates different types of gene dependency relationships (GDRs). Can optionally detect SL where there is a depletion of samples in the bottom left cell of the contingency table (if low-scoring values indicate a reduction or loss of gene function). Performs statistical analysis of depletion with the binomial test. The expected proportion of samples in each contingency table cell is calculated with reference to the size of the clusters.

Usage

mts_patternDetection(
  genepairs = NULL, 
  mixModelClusters1, 
  mixModelClusters2 = NULL, 
  SyntheticLethalityPrediction = TRUE, 
  p_adjustMethod = c("BY","BH"),
  qVal, 
  effectsize = TRUE,  
  effectsize_threshold = NULL, 
  include_reverse_pairs = FALSE,
  directionality = c("depletion", "enrichment"),  
  verbose = FALSE, 
  cores)

Arguments

Value

A data frame with one row per evaluated gene pair, containing the gene names (Gene1, Gene2), the observed count (Actual_Count), the number of cluster combinations, the sample count, the expected count (Expected_Count), and the p_value and false-discovery-rate q_value (plus an Effect_Size column when effectsize = TRUE). An empty data frame with these columns is returned when no gene pair is valid.

See Also

mts_mixModelCluster, mts_mixModelCluster_XPR, mts_crisprPartition, mts_genepairsChunkGeneration, mts_formatMatrix, mts_plotClusterDistribution

Examples

data("mixModelClusters_depMapXPR")

mixturemodelClusters <- mts_mixModelCluster_XPR(dataMatrix = depMapXPR_subset[c(1:12, 20), ])
# partition into two 'clusters' at -0.5
CRISPR_clusters = mts_crisprPartition(dataMatrix = depMapCRISPRscores_subset[1:12, ])

# unidirectional (i.e. symmetrical) evaluation 
# of synthetic lethality depletion relationships with expression data 
predicted_XPR_SL = mts_patternDetection(
  mixModelClusters1=mixModelClusters_depMapXPR,
  qVal=1
)

# unidirectional synthetic lethal depletion prediction with expression data and
# specified gene pairs, using BH FDR correction, filtering using the q-value
# and the effect size 

predicted_XPR_SL2 = mts_patternDetection(
  genepairs = data.frame(Gene1="PSMB8",Gene2="TTC7B"),
  mixModelClusters1 = mixturemodelClusters,
  SyntheticLethalityPrediction = TRUE,
  p_adjustMethod="BH",  
  qVal = 0.05,  effectsize=TRUE)

# bidirectional synthetic lethal depletion prediction with expression and CRISPR data 
# and including reverse pairs 

Predicted_SL3 = mts_patternDetection(
  mixModelClusters1=mixturemodelClusters[1:10], 
  mixModelClusters2=CRISPR_clusters[1:10],
  include_reverse_pairs = FALSE, effectsize=FALSE, qVal =0.05,
  directionality = "enrichment"
)

# tissue specific analysis 
# 1. subset cell lines of interest using mapping file 

Lung_Cancer_cellLines <- depMapTissue_subset$cell_line[
  depMapTissue_subset$tissue == "Lung Cancer"]

# 2. subset XPR mixture model result list 
LungCancer_XPRmixturemodelClusters <- lapply(mixturemodelClusters, function(GMM) {
    GMM[GMM$Sample %in% Lung_Cancer_cellLines, ]
})

# 3. subset CRISPR mixture model result list 
LungCancer_CRISPR_clusters <- lapply(CRISPR_clusters, function(GMM) {
  GMM[GMM$Sample %in% Lung_Cancer_cellLines, ]
})

TissueSpecific_SL = mts_patternDetection(mixModelClusters1 = LungCancer_CRISPR_clusters,
                         mixModelClusters2 = LungCancer_XPRmixturemodelClusters, 
                         include_reverse_pairs = TRUE,
                         SyntheticLethalityPrediction = TRUE,
                         directionality = "enrichment",
                         qVal = 1) 

# bidirectional analysis of all depletion gene dependency relationships 
# for expression and CRISPR data with BY FDR correction 

PredictedGDR = mts_patternDetection(
  mixModelClusters1=mixturemodelClusters[1:10], 
  mixModelClusters2=CRISPR_clusters[1:10],
  SyntheticLethalityPrediction = FALSE,
  directionality = "enrichment",
  p_adjustMethod="BY", effectsize = FALSE, qVal = 0.01
)

Visualise CRISPR and expression data (SynLeGG methods).

Description

Produces a boxplot of CRISPR scores split according to MultiSEp clusters (typically gene expression, but other data types may be used to produce the clusters) and with data points coloured by tissue. Alternatively, may produce a waterfall plot of CRISPR scores coloured by the MultiSEp cluster or a density plot of gene expression values (log2 data are given in the example). In the boxplot, CRISPR scores are split according to the MultiSEp gene expression cluster of the specified mrna gene.

Usage

mts_plotCRISPRGeneCluster(
  mrna_gene, 
  crispr_gene, 
  resultList, 
  crisprMatrix, 
  tissueMatrix, 
  diseaseFilter, 
  mixModelClustersXPR=NULL, 
  plotType
)

Arguments

Value

A ggplot object, which can be printed or saved with ggsave. The plot produced depends on plotType:

• "Integrated": a boxplot of crispr_gene CRISPR scores split according to the MultiSEp gene expression clusters of mrna_gene, with data points coloured by tissue (optionally restricted to a single tissue with diseaseFilter).

• "crispr_only": a waterfall plot of crispr_gene CRISPR scores coloured by the MultiSEp cluster.

• "mrna_only": a density plot of mrna_gene log2 expression values split by MultiSEp cluster.

See Also

mts_mixModelCluster, mts_clusterAvg

Examples

data(depMapXPR_subset)
data(depMapCRISPRscores_subset)
data(depMapTissue_subset)


clusterAssign <- mts_clusterAvg(
  exprsMatrix=depMapXPR_subset[1:5,],
  crisprMatrix=depMapCRISPRscores_subset, 
  cores=1
)

# boxplots
# using clustering results from mts_clusterAvg
mts_plotCRISPRGeneCluster(
  resultList = clusterAssign,
  mrna_gene = "NMT2",
  crispr_gene = "NMT1", 
  crisprMatrix = depMapCRISPRscores_subset, 
  tissueMatrix = depMapTissue_subset,
  plotType = "Integrated"
)

# With clusters defined by mts_mixmodelClusters
clusters_depMapXPR <- mts_mixModelCluster(
    dataMatrix = depMapXPR_subset[1:5,],
    cores = 1
)

mts_plotCRISPRGeneCluster(
  mrna_gene = "DNAJC15", 
  crispr_gene = "DNAJC19", 
  crisprMatrix = depMapCRISPRscores_subset, 
  tissueMatrix = depMapTissue_subset, 
  mixModelClustersXPR = clusters_depMapXPR,
  plotType = "Integrated"
)

# Waterfall plot of CRISPR scores coloured by gene expression cluster

mts_plotCRISPRGeneCluster(
  mrna_gene = "DNAJC15",
  crispr_gene = "DNAJC19",
  crisprMatrix = depMapCRISPRscores_subset,
  tissueMatrix = depMapTissue_subset,
  mixModelClustersXPR = clusters_depMapXPR,
  plotType = "crispr_only"
)

# density plot of gene expression clusters
mts_plotCRISPRGeneCluster(
  resultList = clusterAssign,
  mrna_gene = "NMT2",
  plotType = "mrna_only"
)

# tissue-specific plots
# Boxplot for lung cancer cell lines
mts_plotCRISPRGeneCluster(
  mrna_gene = "DNAJC15",
  crispr_gene = "DNAJC19",
  crisprMatrix = depMapCRISPRscores_subset,
  tissueMatrix = depMapTissue_subset,
  mixModelClustersXPR = clusters_depMapXPR,
  diseaseFilter = 'Lung Cancer',
  plotType = "Integrated"
)

# waterfall plot for lung cancer cell lines
mts_plotCRISPRGeneCluster(
  mrna_gene = "DNAJC15",
  crispr_gene = "DNAJC19",
  crisprMatrix = depMapCRISPRscores_subset,
  tissueMatrix = depMapTissue_subset,
  mixModelClustersXPR = clusters_depMapXPR,
  diseaseFilter = 'Lung Cancer',
  plotType = "crispr_only"
)

Visualise samples and pairwise gene expression clusters

Description

Produces a scatterplot with two selected genes on the x and y axis, the boundaries are determined by the cluster assignments (typically from Gaussian mixture modelling). Density plots of the clusters are also shown for each gene.

Axis scaling is chosen automatically: when a gene has a long high tail (as with many lncRNAs and other wide dynamic-range genes) that axis is square-root, or pseudo-log, transformed for improved visualisation of low-expression clusters (not compressed). This axis scaling is applied independently to each axis, for whichever gene needs it, whether it is supplied as gene1 or gene2.

Usage

mts_plotClusterDistribution(mixModelClusters, 
                            mixModelClusters2 = NULL, 
                            TScluster1 = NULL,
                            TScluster2 = NULL,
                            gene1, gene2, 
                            suppressBlankCanvas = NULL, 
                            gene1_datatype = "", 
                            gene2_datatype = "")

Arguments

Value

A patchwork object combining a scatter plot with the two panels showing density, which can be printed or saved with ggsave. If suppressBlankCanvas is set, the plot is printed and returned invisibly.

See Also

mts_mixModelCluster, mts_mixModelCluster_XPR, mts_crisprPartition, mts_genepairsChunkGeneration, mts_formatMatrix, mts_plotClusterDistribution,

Examples

data("mixModelClusters_depMapXPR")
data("depMapCRISPRscores_subset")
data("depMapTissue_subset")

# Plot with gene expression data for both genes
mts_plotClusterDistribution(
  mixModelClusters = mixModelClusters_depMapXPR, 
  gene1 = "NMT2", 
  gene2="STX2"
)

##
## Plot gene expression with CRISPR gene effect scores
##
# partition CRISPR scores at -0.5
CRISPR_clusters = mts_crisprPartition(dataMatrix = depMapCRISPRscores_subset[4,])
mts_plotClusterDistribution(mixModelClusters = mixModelClusters_depMapXPR,
                            mixModelClusters2 = CRISPR_clusters,
                            gene1 = "NMT2", gene2 = "NMT1",
                            gene1_datatype="log2 gene expression",
                            gene2_datatype = "CRISPR gene effect score")

##
## Tissue-specific plot
##

mixturemodelClusters <- mts_mixModelCluster_XPR(dataMatrix = depMapXPR_subset[14,])
# partition CRISPR scores at -0.5
CRISPR_clusters = mts_crisprPartition(dataMatrix = depMapCRISPRscores_subset[9,])

# 1. subset cell lines of interest using mapping file 
Lung_Cancer_cellLines <- depMapTissue_subset$cell_line[
  depMapTissue_subset$tissue == "Lung Cancer"]

# 2. subset XPR mixture model result list 
LungCancer_XPRmixturemodelClusters <- lapply(mixturemodelClusters, function(GMM) {
    GMM[GMM$Sample %in% Lung_Cancer_cellLines, ]})

# 3. subset CRISPR clusters
LungCancer_CRISPR_Clusters <- lapply(CRISPR_clusters, function(GMM) {
  GMM[GMM$Sample %in% Lung_Cancer_cellLines, ]})

mts_plotClusterDistribution(gene1 = "ACSL1", gene2 = "PSMB5",
                            mixModelClusters = mixturemodelClusters,
                            mixModelClusters2 = CRISPR_clusters,
                            TScluster1 = LungCancer_XPRmixturemodelClusters, 
                            TScluster2 = LungCancer_CRISPR_Clusters)

Visualise clusters for one gene with the mutation status of another gene.

Description

Produces boxplots to visualise mutational status for clusters produced by Gaussian Mixture Modelling, for example derived from transcriptome data.

Usage

mts_plotMutation(resultList=resultList, mrna_gene = mrna_gene, mut_gene = mut_gene, 
mutMatrix = mutMatrix, tissueMatrix = tissueMatrix, diseaseFilter = diseaseFilter)

Arguments

Value

A ggplot object, which can be printed or saved with ggsave. The plot is a boxplot of mrna_gene log2 expression, split by its expression clusters, with data points coloured by the mutation status ("WT" or mutated) of mut_gene; optionally restricted to a single tissue when diseaseFilter is supplied.

See Also

mts_mixModelCluster

Examples

data(depMapXPR_subset)
data(depMapMUT_subset)
data(depMapTissue_subset)

mixModelClusters <- mts_mixModelCluster(dataMatrix = depMapXPR_subset[1:2,])

mts_plotMutation(
  resultList= mixModelClusters,
  mrna_gene = "EIF1AY", 
  mut_gene = "TP53",
  mutMatrix = depMapMUT_subset, 
  tissueMatrix = depMapTissue_subset, 
  diseaseFilter = "Pancreatic Cancer")

Predict the therapeutic population coverage achieved by inhibition of a candidate target gene

Description

This function predicts the therapeutic population coverage achieved by inhibiting each gene in the supplied network of (predicted) synthetic lethal (SL) relationships. A patient with inactivating mutations in partner genes of the candidate target gene is predicted to benefit from inhibition of the target gene because the candidate SL relationship implies that cell death will be caused by inactivation of both genes (one by mutation and one by chemical inhibition). This function can take as input a dataframe of gene pairs (edgeData) or the output from the mts_Mutation() function, both cases also require mutation calls as input.

Usage

  mts_targetCoverageFromMutations(
  mutData, 
  edgeData, 
  rowNum = nrow(edgeData),
  cores
)

Arguments

Value

A data frame containing the following columns:

Examples

data(depMapMUT_subset)
data(mRes)
# An example on the first 5 edges (runs quickly):
TpopCoverage_test <- mts_targetCoverageFromMutations(
  mutData=depMapMUT_subset, 
  edgeData=mRes,
  rowNum=5,
  cores=1
)

Unit Tests for the MultiSEp package

Description

This function allows unit testing to help verify the MultiSEp package integrity, producing a table that summarises the purpose of each test and the outcome (passed/failed).

Usage

mts_tests(mode = c("All", "Demo"),
          cores = 1)

Arguments

Details

The unit tests are designed within the testthat framework and incorporate datasets provided in the MultiSEp package. Run time is approximately 230s using one core (4.8GHz processor). The unit tests cover all of the functions in the MultiSEp package and implement individual expectations such as:

• Expect error: verify error handling is implemented correctly.

• Expect named: ensure code will return a character vector of expected names.

• Expect type: check an object returned by a function is inherited from the expected base type.

• Expect equal: verify code returns an expected reference value.

• Expect length: verify code returns a vector with the specified length.

• Expect no error: check the function runs without returning an error.

Value

A data frame containing the test results as follows:

Test_Case_ID

Identification number of the test case.

Test_Objective

Purpose of the test case.

Outcome

Result of a test case given as "Passed" or "Failed". A "Passed" value indicates to the user that the code performs as expected, "Failed" indicates that it does not.

References

Wickham, H. (2011). 'testthat: Get Started with Testing'. The R Journal (Vol. 3, pp. 5-10). https://journal.r-project.org/archive/2011-1/RJournal_2011-1_Wickham.pdf

Examples

## Demonstration mode

testTab <- mts_tests(mode="Demo")

Example Multisep t-test CRISPR vs XPR results object.

Description

A data frame containing results from multiSEp t-test based analysis of gene expression and CRISPR.

Usage

data("multisepCrisprList")

Format

A data frame with 1428 observations (gene pairs) across 25 variables (columns).

mRNA_gene

Gene id of the gene expression gene

crispr_gene

Gene id of the crispr gene

num_modes

Number of gene expression clusters

nMode1

Number of samples in each gene expression cluster 1

nMode2

Number of samples in each gene expression cluster 2

nMode3

Number of samples in each gene expression cluster 3

nMode4

Number of samples in each gene expression cluster 4

nMode5

Number of samples in each gene expression cluster 5

mMode1

Mean CRISPR score of the crispr gene in each gene expression cluster 1

mMode2

Mean CRISPR score of the crispr gene in each gene expression cluster 2

mMode3

Mean CRISPR score of the crispr gene in each gene expression cluster 3

mMode4

Mean CRISPR score of the crispr gene in each gene expression cluster 4

mMode5

Mean CRISPR score of the crispr gene in each gene expression cluster 5

shift2_1

Fold-change of CRISPR scores between expression modes 2 and 1

shift3_2

Fold-change of CRISPR scores between expression modes 3 and 2

shift4_3

Fold-change of CRISPR scores between expression modes 4 and 3

shift5_4

Fold-change of CRISPR scores between expression modes 5 and 4

pvalue2_1

P-value of CRISPR score difference between expression modes 2 and 1

pvalue3_2

P-value of CRISPR score difference between expression modes 3 and 2

pvalue4_3

P-value of CRISPR score difference between expression modes 4 and 3

pvalue5_4

P-value of CRISPR score difference between expression modes 5 and 4

qvalue2_1

Q-value of CRISPR score difference between expression modes 2 and 1

qvalue3_2

Q-value of CRISPR score difference between expression modes 3 and 2

qvalue4_3

Q-value of CRISPR score difference between expression modes 4 and 3

qvalue5_4

Q-value of CRISPR score difference between expression modes 5 and 4

Source

Derived from analysis of data taken from DepMap 2024Q2 doi:10.25452/figshare.plus.25880521.v1

References

Tsherniak A., Vazquez F., Montgomery P.G., Weir B.A., Kryukov G., Cowley G.S., Gill S., Harrington W.F., Pantel S., Krill-Burger J.M. et al. . Defining a cancer dependency map. Cell. 2017; 170:564-576. doi:10.1016/j.cell.2017.06.010

Examples

data(multisepCrisprList)
multisepCrisprList[1:10,]

Example results from mts_clusterAvg()

Description

mts_clusterAvg() produces a list of sublists corresponding to each gene in the input exprsMatrix object. Each sublist has the following components:

Usage

data("multisepList")

Format

A named list containing ten genes, each gene's element is an unnamed list with two components as follows:

Average CRISPR Scores: The average CRISPR score of each gene in the crisprMatrix object in each cluster assigned by multisep to each gene in the exprsMatrix object.

Sample to Cluster map: The log2 gene expression value for each sample and the associated multisep cluster assignment.

Source

Derived from analysis of DepMap 2024Q2 doi:10.25452/figshare.plus.25880521.v1

References

Tsherniak A., Vazquez F., Montgomery P.G., Weir B.A., Kryukov G., Cowley G.S., Gill S., Harrington W.F., Pantel S., Krill-Burger J.M. et al. . Defining a cancer dependency map. Cell. 2017; 170:564-576. doi:10.1016/j.cell.2017.06.010

See Also

mts_clusterAvg

Examples

data(multisepList)
multisepList[[1]][[2]][1:10,]