Package 'iModMix'

fctAssignmentGenesEnrichr

Description

Enrichr terms for genes (Proteins and transcriptomics)

Usage

fctAssignmentGenesEnrichr(
  clusterAssignmentsProtGenes,
  database = "GO_Biological_Process_2023"
)

Arguments

clusterAssignmentsProtGenes	(data frame containing cluster_assignments and HMDB/Symbol)
database	A list with all the available databases from enrichr

Value

result_list data frame containing cluster_assignments and Enrichr terms

Examples

if (interactive()) {
  # Simulated cluster assignments with gene symbols
  cluster_assignments <- data.frame(
    feature = c("F1", "F2", "F3", "F4"),
    cluster = c("cluster_1", "cluster_1", "cluster_2", "cluster_2"),
    col = c("#FF0000", "#FF0000", "#00FF00", "#00FF00"),
    feature_name = c("TP53", "BRCA1", "EGFR", "MYC"),
    stringsAsFactors = FALSE
  )

  # Run enrichment analysis using Enrichr (requires internet connection)
  enriched_results <- fctAssignmentGenesEnrichr(
    clusterAssignmentsProtGenes = cluster_assignments,
    database = "GO_Biological_Process_2023"
  )

  # View results
  head(enriched_results)
}

---

fctClusterAssignments

Description

Cluster assignments with annotation.

Usage

fctClusterAssignments(cluster, phenoData = NULL, selectedColumns = NULL)

Arguments

cluster	A data frame containing cluster assignments.
phenoData	A data frame with the annotation names of the Dataset. Should have a column called Feature_ID.
selectedColumns	Columns selected for the user to merge in the cluster assignments table

Value

A data frame containing cluster assignments and selected columns by user.

Examples

# Simulated cluster assignment data
cluster_df <- data.frame(
  feature = paste0("F", 1:5),
  cluster = paste0("cluster_", 1:5),
  col = RColorBrewer::brewer.pal(5, "Set1"),
  stringsAsFactors = FALSE
)

# Simulated annotation data
pheno_df <- data.frame(
  Feature_ID = paste0("F", 1:5),
  Symbol = c("GeneA", "GeneB", "GeneC", "GeneD", "GeneE"),
  Description = paste("Description", 1:5),
  stringsAsFactors = FALSE
)

# Run without selected columns
fctClusterAssignments(cluster_df, phenoData = pheno_df)

# Run with selected columns
fctClusterAssignments(cluster_df, phenoData = pheno_df, selectedColumns = "Description")

---

fctEigengenes

Description

Calculates module eigengenes using the WGCNA package.

Usage

fctEigengenes(loadData = loadData, clusterAssignments = clusterAssignments)

Arguments

loadData	A prepossessed data matrix resulting from "load data" function
clusterAssignments	A vector of cluster assignments for each feature.

Value

A list containing:

module_eigenmetab_List_Me	The full result from WGCNA::moduleEigengenes.
module_eigenmetab_Me	The matrix of module eigengenes.
feature_mat_t	The filtered and scaled feature matrix.

Examples

# Simulate a small expression matrix
set.seed(123)
mat <- matrix(rnorm(100), nrow = 10, ncol = 10)
colnames(mat) <- paste0("Gene", 1:10)
rownames(mat) <- paste0("Sample", 1:10)

# Simulate cluster assignments
clusters <- rep(1:2, each = 5)

# Run Eigengenes function
result <- fctEigengenes(loadData = mat, clusterAssignments = clusters)

# View eigengene matrix
head(result$module_eigenmetab_Me)

---

fctFeaturesAnnotCorrelation

Description

Calculate correlation between the features of top correlated modules.

Usage

fctFeaturesAnnotCorrelation(
  corDataiDataj,
  clusterAssignmentsD1,
  clusterAssignmentsD2,
  loadData1 = loadData1,
  loadData2 = loadData2,
  topn = 5
)

Arguments

corDataiDataj	A data frame with the first principal component of each protein cluster and their correlations.
clusterAssignmentsD1	A data frame containing cluster assignments for metabolites.
clusterAssignmentsD2	A data frame containing cluster assignments and Enrichr terms for proteins.
loadData1	A prepossessed data matrix resulting from "load data" function from Data1.
loadData2	A prepossessed data matrix resulting from "load data" function from Data2.
topn	The number of top correlations to select.

Value

A list containing important features and correlation matrices.

Examples

# Simulated correlation data
corDataiDataj <- data.frame(
  from = c("blue", "green"),
  to = c("red", "yellow"),
  value = c(0.9, 0.85),
  stringsAsFactors = FALSE
)

# Simulated cluster assignments
clusterAssignmentsD1 <- data.frame(
  feature = c("M1", "M2"),
  col = c("blue", "green"),
  stringsAsFactors = FALSE
)
clusterAssignmentsD2 <- data.frame(
  feature = c("P1", "P2"),
  col = c("red", "yellow"),
  stringsAsFactors = FALSE
)

# Simulated expression matrices
loadData1 <- matrix(rnorm(20), nrow = 5, ncol = 4)
colnames(loadData1) <- c("M1", "M2", "X1", "X2")
rownames(loadData1) <- paste0("Sample", 1:5)

loadData2 <- matrix(rnorm(20), nrow = 5, ncol = 4)
colnames(loadData2) <- c("P1", "P2", "Y1", "Y2")
rownames(loadData2) <- paste0("Sample", 1:5)

# Run the function
result <- fctFeaturesAnnotCorrelation(
  corDataiDataj = corDataiDataj,
  clusterAssignmentsD1 = clusterAssignmentsD1,
  clusterAssignmentsD2 = clusterAssignmentsD2,
  loadData1 = loadData1,
  loadData2 = loadData2,
  topn = 2
)

# View top correlations
result$Top_correlations

---

fctHierarchicalCluster

Description

Perform hierarchical clustering of a partial correlation matrix. By default uses the topological overlap measure (TOM)

Usage

fctHierarchicalCluster(parcorMat, tom = TRUE, minModuleSize = 10)

Arguments

parcorMat	A partial correlation matrix from glassoFast$wi: Estimated inverse covariance matrix
tom	TRUE: topological overlap measure; FALSE: partial correlation matrix; default is TRUE.
minModuleSize	Smallest modules to be generated; default is 10

Value

result_list List containing TOM_diss (TOM dissimilarity; NA if argument TOM = FALSE), hclustTree (hierarchical clustering object), dynamicMods (index of module membership for features), and mod_count (number of modules).

Examples

# Simulate a small partial correlation matrix
set.seed(123)
mat <- matrix(rnorm(100), nrow = 10)
colnames(mat) <- paste0("Gene", 1:10)
rownames(mat) <- paste0("Sample", 1:10)

# Compute covariance and partial correlation matrix
cov_mat <- cov(mat)
parcorMat <- glassoFast::glassoFast(cov_mat, rho = 0.25)$wi
diag(parcorMat) <- NA
colnames(parcorMat) <- rownames(parcorMat) <- paste0("Gene", 1:10)

# Run hierarchical clustering
result <- fctHierarchicalCluster(parcorMat, tom = TRUE, minModuleSize = 3)

# View module assignments
head(result$cluster_assignments)

---

fctiModMix2SE

Description

Converts expression data and corresponding sample metadata into a Bioconductor-compliant SummarizedExperiment object.

Usage

fctiModMix2SE(dataExp, metadata)

Arguments

dataExp	A data frame with metabolites (features) in rows and samples in columns. The first column should contain feature IDs (e.g., metabolite names).
metadata	A data frame containing sample metadata. Must include a column named "Sample" that matches the column names of dataExp (excluding the first column).

Value

A SummarizedExperiment object.

Examples

dataExp <- data.frame(Feature_ID = paste0("M", 1:5),
                      S1 = rnorm(5), S2 = rnorm(5))
metadata <- data.frame(Sample = c("S1", "S2"),
                       Group = c("A", "B"))
se <- fctiModMix2SE(dataExp, metadata)
se

---

fctLoadData

Description

A function to load and preprocess the expression data.

Usage

fctLoadData(expressionMat = expressionMat)

Arguments

expressionMat

A feature matrix (e.g., gene expression) with samples in rows and features (e.g., genes) in columns. Row names must be unique.

Value

A prepossessed data matrix.

Examples

# Simulated expression matrix with missing values
set.seed(123)
mat <- matrix(rnorm(1000), nrow = 10, ncol = 100)
colnames(mat) <- paste0("Gene", 1:100)
rownames(mat) <- paste0("Sample", 1:10)
mat[sample(length(mat), 50)] <- NA  # introduce some NAs

# Add a Feature_ID column to mimic expected input
expressionMat <- cbind(Feature_ID = paste0("F", 1:10), mat)
expressionMat <- as.data.frame(expressionMat)

# Convert Feature_ID to character (if needed)
expressionMat$Feature_ID <- as.character(expressionMat$Feature_ID)

# Run the function
processed_data <- fctLoadData(expressionMat)

---

fctModulesCorrelation

Description

Calculates correlations between omic modules.

Usage

fctModulesCorrelation(eigengenesList, clusterList, threshold = 0.5)

Arguments

eigengenesList	List with the first principal component of each cluster.
clusterList	list containing cluster_assignments matrix.
threshold	A numeric value to filter correlations. Default is 0.5.

Value

Return the Top_correlations between Prot and metab eigengenes, graph of red, correlation plot.

Examples

# Simulate eigengene matrices for two omics datasets
set.seed(123)
eig1 <- matrix(rnorm(50), nrow = 10, ncol = 5)
colnames(eig1) <- paste0("ME", 1:5)
eig2 <- matrix(rnorm(50), nrow = 10, ncol = 5)
colnames(eig2) <- paste0("ME", 1:5)

# Simulate cluster assignments for each omic
cluster1 <- data.frame(
  feature = paste0("F", 1:5),
  cluster = paste0("cluster_", 1:5),
  col = RColorBrewer::brewer.pal(5, "Set1"),
  stringsAsFactors = FALSE
)
cluster2 <- data.frame(
  feature = paste0("F", 6:10),
  cluster = paste0("cluster_", 6:10),
  col = RColorBrewer::brewer.pal(5, "Set2"),
  stringsAsFactors = FALSE
)

# Run the correlation function
result <- fctModulesCorrelation(
  eigengenesList = list(eig1, eig2),
  clusterList = list(cluster1, cluster2),
  threshold = 0.5
)

# View top correlations
head(result$Top_cor_Prot_metab)

---

fctPartialCors

Description

Calculates partial correlations based on graphical lasso (https://www.rdocumentation.org/packages/glassoFast/versions/1.0/topics/glassoFast).

Usage

fctPartialCors(loadData, rho = 0.25)

Arguments

loadData	A prepossessed data matrix resulting from "load data" function.
rho	The regularization parameter for lasso. (a non-negative value or a p by p matrix of regularization parameters).

Value

partial_cor_mat A partial correlation matrix with NA's in the diagonal.

Examples

# Simulate a small expression matrix
set.seed(123)
mat <- matrix(rnorm(100), nrow = 10, ncol = 10)
colnames(mat) <- paste0("Gene", 1:10)
rownames(mat) <- paste0("Sample", 1:10)

# Run partial correlation
partial_cor_matrix <- fctPartialCors(mat, rho = 0.25)

# View a portion of the result
partial_cor_matrix[1:5, 1:5]

---

fctPerformClassification

Description

Performs classification using different methods such as Welch’s T-test, Random Forest, and K-Nearest Neighbors.

Usage

fctPerformClassification(
  eigengeneData,
  metadata,
  phenotypeVariable,
  significanceThreshold = 0.05
)

Arguments

eigengeneData	A data frame of eigengenes organized with patient IDs in rows and variables in columns.
metadata	A data frame containing metadata, with a column call "Sample" that matches patient IDs in eigengeneData.
phenotypeVariable	The variable selected by the user in the Shiny app (response variable).
significanceThreshold	A numeric value to filter p-value. Default is 00.5.

Value

A data frame with metrics such as AUC, Accuracy, and Error Rate for each binary classification.

Examples

# Simulate eigengene data
set.seed(123)
eigengeneData <- as.data.frame(matrix(rnorm(100), nrow = 10, ncol = 10))
colnames(eigengeneData) <- paste0("ME", 1:10)
rownames(eigengeneData) <- paste0("Sample", 1:10)
# Simulate metadata with a binary phenotype
metadata <- data.frame(
  Sample = paste0("Sample", 1:10),
  Phenotype = rep(c("A", "B"), each = 5),
  stringsAsFactors = FALSE
)
# Run classification
result <- fctPerformClassification(
  eigengeneData = eigengeneData,
  metadata = metadata,
  phenotypeVariable = "Phenotype",
  significanceThreshold = 0.05
)

# View results
head(result$result)

---

Run the Shiny Application

Description

Run the Shiny Application

Usage

run_app(
  onStart = NULL,
  options = list(),
  enableBookmarking = NULL,
  uiPattern = "/",
  ...
)

Arguments

onStart	A function that will be called before the app is actually run. This is only needed for shinyAppObj, since in the shinyAppDir case, a global.R file can be used for this purpose.
options	Named options that should be passed to the runApp call (these can be any of the following: "port", "launch.browser", "host", "quiet", "display.mode" and "test.mode"). You can also specify width and height parameters which provide a hint to the embedding environment about the ideal height/width for the app.
enableBookmarking	Can be one of "url", "server", or "disable". The default value, NULL, will respect the setting from any previous calls to enableBookmarking(). See enableBookmarking() for more information on bookmarking your app.
uiPattern	A regular expression that will be applied to each GET request to determine whether the ui should be used to handle the request. Note that the entire request path must match the regular expression in order for the match to be considered successful.
...	arguments to pass to golem_opts. See '?golem::get_golem_options' for more details.

Value

A Shiny application object that launches the iModMix interface.

---