Package 'MOFSR'

Description

Factor analysis methods for multi-omics data integration.

Performs multi-view factor analysis for multi-omics integration. This is a simplified Bayesian factor model with ARD priors for sparsity. For the full MOFA implementation, use the MOFA2 Bioconductor package.

Usage

multi_view_factor_analysis(
  data_list,
  n_factors = 10,
  max_iter = 1000,
  tol = 1e-05,
  sparsity_prior = TRUE,
  verbose = FALSE
)

Arguments

Details

This function implements a simplified multi-view factor model: Y_m = W_m * Z^T + E_m, where Y_m is the data for view m, W_m are the loadings, Z are the latent factors shared across views, and E_m is Gaussian noise. ARD (Automatic Relevance Determination) priors are used for automatic factor pruning.

Value

List with factors (Z), weights (W), variance explained, and convergence info.

Note

For production use of MOFA, we recommend the MOFA2 Bioconductor package which provides the full implementation with Python backend.

Author(s)

Zaoqu Liu; Email: [email protected]

Description

Pure R implementation of iClusterBayes for multi-omics clustering.

Performs Bayesian integrative clustering using Gibbs sampling.

Usage

icluster_bayes(
  data_list,
  k,
  n_burnin = 1000,
  n_sample = 2000,
  n_thin = 1,
  prior_alpha = 1,
  prior_sigma = c(1, 1),
  verbose = FALSE
)

Arguments

Value

List with cluster assignments, posterior matrices, and parameters.

Author(s)

Zaoqu Liu

References

Mo Q, et al. A fully Bayesian latent variable model for integrative clustering analysis of multi-type omics data. Biostatistics. 2018.

Description

Internal implementation of NEMO algorithm for multi-omics clustering.

Usage

.NEMO_NEIGHBORS_RATIO

Format

An object of class numeric of length 1.

Author(s)

Zaoqu Liu

References

Rappoport N, Shamir R. NEMO: cancer subtyping by integration of partial multi-omic data. Bioinformatics. 2019.

Description

Internal implementation of SNF algorithm for multi-omics data integration.

Computes an affinity matrix using local Gaussian kernel.

Usage

snf_affinity_matrix(diff, K = 20, sigma = 0.5)

Arguments

Value

Affinity matrix with exponential similarity.

Author(s)

Zaoqu Liu

References

Wang B, et al. Similarity Network Fusion for Aggregating Data Types on a Genomic Scale. Nat Methods. 2014;11(3):333-337.

Wang B, et al. Nat Methods. 2014 - Equation in Methods section

Description

Aligns samples across multiple datasets to common samples.

Usage

align_samples(data_list)

Arguments

Value

List of aligned matrices with common samples.

Description

Performs Bayesian Consensus Clustering.

Usage

bcc_cluster(
  data_list,
  k,
  n_iter = 1000,
  n_burnin = 500,
  alpha = 1,
  verbose = FALSE
)

Arguments

Value

List with cluster assignments and consensus matrix.

Description

Faster version of BCC using EM-like approach.

Usage

bcc_cluster_fast(data_list, k, max_iter = 50)

Arguments

Value

List with cluster assignments.

Description

Computes CH index to evaluate clustering quality.

Usage

calc_chi(
  hclust_result,
  dist_matrix = NULL,
  max_clusters = round(1 + 3.3 * log10(length(hclust_result$order)))
)

Arguments

Value

Vector of CH index values.

Description

Computes PAC to evaluate clustering stability.

Usage

calc_pac(consensus_result, range_clusters = 2:6, x1 = 0.1, x2 = 0.9)

Arguments

Value

Data frame with PAC values for each K.

Description

Calculates the Calinski-Harabasz index for evaluating clustering quality.

Usage

CalCHI(
  hclust_result,
  dist_matrix = NULL,
  max_clusters = round(1 + 3.3 * log10(length(hclust_result$order)))
)

Arguments

Value

A vector containing CH index values for each number of clusters.

Author(s)

Zaoqu Liu; Email: [email protected]

Description

This function calculates the Proportion of Ambiguous Clustering (PAC) to help evaluate the optimal number of clusters in a consensus clustering analysis.

Usage

CalPAC(consensus_result, range_clusters = 2:6, x1 = 0.1, x2 = 0.9)

Arguments

Details

The PAC is calculated as the difference between the cumulative distribution function values at two thresholds, x2 and x1, on the consensus matrix values. A lower PAC indicates a more stable clustering solution.

Value

A data frame containing the PAC values for each number of clusters.

Author(s)

Zaoqu Liu; Email: [email protected]

Examples

data <- mtcars
cc_res <- RunCC(data)
pac_values <- CalPAC(cc_res)
pac_values

Description

Checks if samples are aligned across datasets.

Usage

check_sample_alignment(data_list, strict = FALSE)

Arguments

Value

TRUE if aligned, otherwise prints discrepancies.

Description

Performs multi-kernel learning based clustering.

Usage

cimlr_cluster(data_list, k, n_kernels = 10, max_iter = 30)

Arguments

Value

List with clusters, kernel, weights.

Description

Ranks features based on their contribution to clustering.

Usage

cimlr_feature_ranking(data_list, cluster)

Arguments

Value

List of feature importance for each data type.

Description

This function performs classification using AdaBoost to predict cluster assignments for test data based on trained models from training data and cluster markers.

Usage

Classifier.Adaboost(
  data.test,
  data.train,
  cluster.data,
  cluster.markers,
  scale = TRUE
)

Arguments

Details

The function operates as follows: 1. Ensures that the 'cluster.data' has the correct column names. 2. Adds a one-hot encoded matrix for cluster assignments. 3. Scales the test data for prediction. 4. Selects genes that are common between the test and training datasets. 5. Uses glmnet to identify the important markers for each cluster and trains an AdaBoost model for classification. 6. Predicts the cluster for test samples and provides probabilities for each cluster.

Value

A data frame with: - ID: The sample identifier. - Probabilities: The probabilities for each cluster assignment. - Predict: The predicted cluster label for each sample.

Author(s)

Zaoqu Liu; Email: [email protected]

Examples

cluster.data <- data.frame(
  Sample = paste0("Sample", 1:60),
  Cluster = rep(paste0("C", 1:3), each = 20)
)
data.train <- matrix(rnorm(6000),
  nrow = 100,
  dimnames = list(paste0("Gene", 1:100), cluster.data$Sample)
)
data.test <- matrix(rnorm(5000),
  nrow = 100,
  dimnames = list(paste0("Gene", 1:100), paste0("P", 1:50))
)
cluster.markers <- setNames(
  lapply(
    unique(cluster.data$Cluster),
    function(cluster) {
      data.frame(Gene = sample(rownames(data.train), 10))
    }
  ),
  unique(cluster.data$Cluster)
)
result <- Classifier.Adaboost(
  data.test = data.test, data.train = data.train,
  cluster.data, cluster.markers
)
head(result)

Description

This function performs classification using Decision Tree to predict cluster assignments for test data based on trained models from training data and cluster markers.

Usage

Classifier.DT(
  data.test,
  data.train,
  cluster.data,
  cluster.markers,
  scale = TRUE
)

Arguments

Details

The function operates as follows: 1. Ensures that the 'cluster.data' has the correct column names. 2. Adds a one-hot encoded matrix for cluster assignments. 3. Scales the test data for prediction. 4. Selects genes that are common between the test and training datasets. 5. Uses glmnet to identify the important markers for each cluster and trains a Decision Tree model for classification. 6. Predicts the cluster for test samples and provides probabilities for each cluster.

Value

A data frame with: - ID: The sample identifier. - Probabilities: The probabilities for each cluster assignment. - Predict: The predicted cluster label for each sample.

Author(s)

Zaoqu Liu; Email: [email protected]

Examples

cluster.data <- data.frame(
  Sample = paste0("Sample", 1:60),
  Cluster = rep(paste0("C", 1:3), each = 20)
)
data.train <- matrix(rnorm(6000),
  nrow = 100,
  dimnames = list(paste0("Gene", 1:100), cluster.data$Sample)
)
data.test <- matrix(rnorm(5000),
  nrow = 100,
  dimnames = list(paste0("Gene", 1:100), paste0("P", 1:50))
)
cluster.markers <- setNames(
  lapply(
    unique(cluster.data$Cluster),
    function(cluster) {
      data.frame(Gene = sample(rownames(data.train), 10))
    }
  ),
  unique(cluster.data$Cluster)
)
result <- Classifier.DT(
  data.test = data.test, data.train = data.train,
  cluster.data, cluster.markers
)
head(result)

Description

This function performs classification using Elastic Net to predict cluster assignments for test data based on trained models from training data and cluster markers.

Usage

Classifier.Enet(
  data.test,
  data.train,
  cluster.data,
  cluster.markers,
  scale = TRUE
)

Arguments

Details

The function operates as follows: 1. Ensures that the 'cluster.data' has the correct column names. 2. Adds a one-hot encoded matrix for cluster assignments. 3. Scales the test data for prediction. 4. Selects genes that are common between the test and training datasets. 5. Uses glmnet to identify the important markers for each cluster and trains an Elastic Net model for classification. 6. Predicts the cluster for test samples and provides probabilities for each cluster.

Value

A data frame with: - ID: The sample identifier. - Probabilities: The probabilities for each cluster assignment. - Predict: The predicted cluster label for each sample.

Author(s)

Zaoqu Liu; Email: [email protected]

Examples

cluster.data <- data.frame(
  Sample = paste0("Sample", 1:60),
  Cluster = rep(paste0("C", 1:3), each = 20)
)
data.train <- matrix(rnorm(6000),
  nrow = 100,
  dimnames = list(paste0("Gene", 1:100), cluster.data$Sample)
)
data.test <- matrix(rnorm(5000),
  nrow = 100,
  dimnames = list(paste0("Gene", 1:100), paste0("P", 1:50))
)
cluster.markers <- setNames(
  lapply(
    unique(cluster.data$Cluster),
    function(cluster) {
      data.frame(Gene = sample(rownames(data.train), 10))
    }
  ),
  unique(cluster.data$Cluster)
)
result <- Classifier.Enet(
  data.test = data.test, data.train = data.train,
  cluster.data, cluster.markers
)
head(result)

Description

This function performs classification using pathway enrichment analysis and a neural network to predict cluster assignments for test data based on trained models from training data and cluster markers.

Usage

Classifier.Enrichment(
  data.test,
  data.train,
  cluster.data,
  cluster.markers,
  scale = TRUE,
  nCores = 5
)

Arguments

Details

The function operates as follows: 1. Ensures that the 'cluster.data' has the correct column names. 2. Adds a one-hot encoded matrix for cluster assignments. 3. Scales the test data for prediction. 4. Selects genes that are common between the test and training datasets. 5. Filters out markers with low odds ratio. 6. Performs pathway enrichment analysis using the ssMwwGST method.

Value

A data frame with: - ID: The sample identifier. - Cluster: The predicted cluster label for each sample. - NES: Normalized Enrichment Score (NES) for each cluster assignment.

Author(s)

Zaoqu Liu; Email: [email protected]

Examples

cluster.data <- data.frame(
  Sample = paste0("Sample", 1:60),
  Cluster = rep(paste0("C", 1:3), each = 20)
)
data.train <- matrix(rnorm(6000),
  nrow = 100,
  dimnames = list(paste0("Gene", 1:100), cluster.data$Sample)
)
data.test <- matrix(rnorm(5000),
  nrow = 100,
  dimnames = list(paste0("Gene", 1:100), paste0("P", 1:50))
)
cluster.markers <- setNames(
  lapply(
    unique(cluster.data$Cluster),
    function(cluster) {
      data.frame(Gene = sample(rownames(data.train), 10))
    }
  ),
  unique(cluster.data$Cluster)
)
result <- Classifier.Enrichment(
  data.test = data.test, data.train = data.train,
  cluster.data, cluster.markers
)
head(result)

Description

This function performs classification using Gradient Boosted Decision Trees (GBDT) to predict cluster assignments for test data based on trained models from training data and cluster markers.

Usage

Classifier.GBDT(
  data.test,
  data.train,
  cluster.data,
  cluster.markers,
  scale = TRUE
)

Arguments

Details

The function operates as follows: 1. Ensures that the 'cluster.data' has the correct column names. 2. Adds a one-hot encoded matrix for cluster assignments. 3. Scales the test data for prediction. 4. Selects genes that are common between the test and training datasets. 5. Uses glmnet to identify the important markers for each cluster and trains a Gradient Boosted Decision Trees model for classification. 6. Predicts the cluster for test samples and provides probabilities for each cluster.

Value

A data frame with: - ID: The sample identifier. - Probabilities: The probabilities for each cluster assignment. - Predict: The predicted cluster label for each sample.

Author(s)

Zaoqu Liu; Email: [email protected]

Examples

cluster.data <- data.frame(
  Sample = paste0("Sample", 1:60),
  Cluster = rep(paste0("C", 1:3), each = 20)
)
data.train <- matrix(rnorm(6000),
  nrow = 100,
  dimnames = list(paste0("Gene", 1:100), cluster.data$Sample)
)
data.test <- matrix(rnorm(5000),
  nrow = 100,
  dimnames = list(paste0("Gene", 1:100), paste0("P", 1:50))
)
cluster.markers <- setNames(
  lapply(
    unique(cluster.data$Cluster),
    function(cluster) {
      data.frame(Gene = sample(rownames(data.train), 10))
    }
  ),
  unique(cluster.data$Cluster)
)
result <- Classifier.GBDT(
  data.test = data.test, data.train = data.train,
  cluster.data, cluster.markers
)
head(result)

Description

This function performs classification using k-Nearest Neighbors to predict cluster assignments for test data based on trained models from training data and cluster markers.

Usage

Classifier.kNN(
  data.test,
  data.train,
  cluster.data,
  cluster.markers,
  scale = TRUE
)

Arguments

Details

The function operates as follows: 1. Ensures that the 'cluster.data' has the correct column names and matches training data. 2. Scales the test data for prediction if 'scale' is TRUE. 3. Selects genes that are common between the test and training datasets. 4. Uses glmnet to identify the important markers for each cluster and trains a k-Nearest Neighbors (kNN) model for classification. 5. Predicts the cluster for test samples and provides probabilities for each cluster.

Value

A data frame with: - ID: The sample identifier. - Cluster: The predicted cluster label for each sample. - Probabilities: The probabilities for each cluster assignment.

Author(s)

Zaoqu Liu; Email: [email protected]

Examples

cluster.data <- data.frame(
  Sample = paste0("Sample", 1:60),
  Cluster = rep(paste0("C", 1:3), each = 20)
)
data.train <- matrix(rnorm(6000),
  nrow = 100,
  dimnames = list(paste0("Gene", 1:100), cluster.data$Sample)
)
data.test <- matrix(rnorm(5000),
  nrow = 100,
  dimnames = list(paste0("Gene", 1:100), paste0("P", 1:50))
)
cluster.markers <- setNames(
  lapply(
    unique(cluster.data$Cluster),
    function(cluster) {
      data.frame(Gene = sample(rownames(data.train), 10))
    }
  ),
  unique(cluster.data$Cluster)
)
result <- Classifier.kNN(
  data.test = data.test, data.train = data.train,
  cluster.data, cluster.markers
)
head(result)

Description

This function performs classification using LASSO (Least Absolute Shrinkage and Selection Operator) to predict cluster assignments for test data based on trained models from training data and cluster markers.

Usage

Classifier.LASSO(
  data.test,
  data.train,
  cluster.data,
  cluster.markers,
  scale = TRUE
)

Arguments

Details

The function operates as follows: 1. Ensures that the 'cluster.data' has the correct column names. 2. Adds a one-hot encoded matrix for cluster assignments. 3. Scales the test data for prediction. 4. Selects genes that are common between the test and training datasets. 5. Uses glmnet to identify the important markers for each cluster and trains a LASSO model for classification. 6. Predicts the cluster for test samples and provides probabilities for each cluster.

Value

A data frame with: - ID: The sample identifier. - Probabilities: The probabilities for each cluster assignment. - Predict: The predicted cluster label for each sample.

Author(s)

Zaoqu Liu; Email: [email protected]

Examples

cluster.data <- data.frame(
  Sample = paste0("Sample", 1:60),
  Cluster = rep(paste0("C", 1:3), each = 20)
)
data.train <- matrix(rnorm(6000),
  nrow = 100,
  dimnames = list(paste0("Gene", 1:100), cluster.data$Sample)
)
data.test <- matrix(rnorm(5000),
  nrow = 100,
  dimnames = list(paste0("Gene", 1:100), paste0("P", 1:50))
)
cluster.markers <- setNames(
  lapply(
    unique(cluster.data$Cluster),
    function(cluster) {
      data.frame(Gene = sample(rownames(data.train), 10))
    }
  ),
  unique(cluster.data$Cluster)
)
result <- Classifier.LASSO(
  data.test = data.test, data.train = data.train,
  cluster.data, cluster.markers
)
head(result)

Description

This function performs classification using Linear Discriminant Analysis (LDA) to predict cluster assignments for test data based on trained models from training data and cluster markers.

Usage

Classifier.LDA(
  data.test,
  data.train,
  cluster.data,
  cluster.markers,
  scale = TRUE
)

Arguments

Details

The function operates as follows: 1. Ensures that the 'cluster.data' has the correct column names. 2. Adds a one-hot encoded matrix for cluster assignments. 3. Scales the test data for prediction. 4. Selects genes that are common between the test and training datasets. 5. Uses glmnet to identify the important markers for each cluster and trains an LDA model for classification. 6. Predicts the cluster for test samples and provides probabilities for each cluster.

Value

A data frame with: - ID: The sample identifier. - Probabilities: The probabilities for each cluster assignment. - Predict: The predicted cluster label for each sample.

Author(s)

Zaoqu Liu; Email: [email protected]

Examples

cluster.data <- data.frame(
  Sample = paste0("Sample", 1:60),
  Cluster = rep(paste0("C", 1:3), each = 20)
)
data.train <- matrix(rnorm(6000),
  nrow = 100,
  dimnames = list(paste0("Gene", 1:100), cluster.data$Sample)
)
data.test <- matrix(rnorm(5000),
  nrow = 100,
  dimnames = list(paste0("Gene", 1:100), paste0("P", 1:50))
)
cluster.markers <- setNames(
  lapply(
    unique(cluster.data$Cluster),
    function(cluster) {
      data.frame(Gene = sample(rownames(data.train), 10))
    }
  ),
  unique(cluster.data$Cluster)
)
result <- Classifier.LDA(
  data.test = data.test, data.train = data.train,
  cluster.data, cluster.markers
)
head(result)

Description

This function performs classification using Naive Bayes to predict cluster assignments for test data based on trained models from training data and cluster markers.

Usage

Classifier.NBayes(
  data.test,
  data.train,
  cluster.data,
  cluster.markers,
  scale = TRUE
)

Arguments

Details

The function operates as follows: 1. Ensures that the 'cluster.data' has the correct column names. 2. Adds a one-hot encoded matrix for cluster assignments. 3. Scales the test data for prediction. 4. Selects genes that are common between the test and training datasets. 5. Uses glmnet to identify the important markers for each cluster and trains a Naive Bayes model for classification. 6. Predicts the cluster for test samples and provides probabilities for each cluster.

Value

A data frame with: - ID: The sample identifier. - Probabilities: The probabilities for each cluster assignment. - Predict: The predicted cluster label for each sample.

Author(s)

Zaoqu Liu; Email: [email protected]

Examples

cluster.data <- data.frame(
  Sample = paste0("Sample", 1:60),
  Cluster = rep(paste0("C", 1:3), each = 20)
)
data.train <- matrix(rnorm(6000),
  nrow = 100,
  dimnames = list(paste0("Gene", 1:100), cluster.data$Sample)
)
data.test <- matrix(rnorm(5000),
  nrow = 100,
  dimnames = list(paste0("Gene", 1:100), paste0("P", 1:50))
)
cluster.markers <- setNames(
  lapply(
    unique(cluster.data$Cluster),
    function(cluster) {
      data.frame(Gene = sample(rownames(data.train), 10))
    }
  ),
  unique(cluster.data$Cluster)
)
result <- Classifier.NBayes(
  data.test = data.test, data.train = data.train,
  cluster.data, cluster.markers
)
head(result)

Description

This function performs classification using a neural network to predict cluster assignments for test data based on trained models from training data and cluster markers.

Usage

Classifier.NNet(
  data.test,
  data.train,
  cluster.data,
  cluster.markers,
  scale = TRUE
)

Arguments

Details

The function operates as follows: 1. Ensures that the 'cluster.data' has the correct column names. 2. Adds a one-hot encoded matrix for cluster assignments. 3. Scales the test data for prediction. 4. Selects genes that are common between the test and training datasets. 5. Uses glmnet to identify the important markers for each cluster and trains a neural network model for classification. 6. Predicts the cluster for test samples and provides probabilities for each cluster.

Value

A data frame with: - ID: The sample identifier. - Probabilities: The probabilities for each cluster assignment. - Predict: The predicted cluster label for each sample.

Author(s)

Zaoqu Liu; Email: [email protected]

Examples

cluster.data <- data.frame(
  Sample = paste0("Sample", 1:60),
  Cluster = rep(paste0("C", 1:3), each = 20)
)
data.train <- matrix(rnorm(6000),
  nrow = 100,
  dimnames = list(paste0("Gene", 1:100), cluster.data$Sample)
)
data.test <- matrix(rnorm(5000),
  nrow = 100,
  dimnames = list(paste0("Gene", 1:100), paste0("P", 1:50))
)
cluster.markers <- setNames(
  lapply(
    unique(cluster.data$Cluster),
    function(cluster) {
      data.frame(Gene = sample(rownames(data.train), 10))
    }
  ),
  unique(cluster.data$Cluster)
)
result <- Classifier.NNet(
  data.test = data.test, data.train = data.train,
  cluster.data, cluster.markers
)
head(result)

Description

This function performs classification using PCA and a neural network to predict cluster assignments for test data based on trained models from training data and cluster markers.

Usage

Classifier.PCA(
  data.test,
  data.train,
  cluster.data,
  cluster.markers,
  scale = TRUE
)

Arguments

Details

The function operates as follows: 1. Ensures that the 'cluster.data' has the correct column names. 2. Adds a one-hot encoded matrix for cluster assignments. 3. Scales the test data for prediction. 4. Selects genes that are common between the test and training datasets. 5. Uses glmnet to identify the important markers for each cluster and performs PCA to reduce dimensionality. 6. Trains a neural network model for classification using the top PCs. 7. Predicts the cluster for test samples and provides probabilities for each cluster.

Value

A data frame with: - ID: The sample identifier. - Probabilities: The probabilities for each cluster assignment. - Predict: The predicted cluster label for each sample.

Author(s)

Zaoqu Liu; Email: [email protected]

Examples

cluster.data <- data.frame(
  Sample = paste0("Sample", 1:60),
  Cluster = rep(paste0("C", 1:3), each = 20)
)
data.train <- matrix(rnorm(6000),
  nrow = 100,
  dimnames = list(paste0("Gene", 1:100), cluster.data$Sample)
)
data.test <- matrix(rnorm(5000),
  nrow = 100,
  dimnames = list(paste0("Gene", 1:100), paste0("P", 1:50))
)
cluster.markers <- setNames(
  lapply(
    unique(cluster.data$Cluster),
    function(cluster) {
      data.frame(Gene = sample(rownames(data.train), 10))
    }
  ),
  unique(cluster.data$Cluster)
)
result <- Classifier.PCA(
  data.test = data.test, data.train = data.train,
  cluster.data, cluster.markers
)
head(result)

Description

This function performs classification using Random Forest to predict cluster assignments for test data based on trained models from training data and cluster markers.

Usage

Classifier.RF(
  data.test,
  data.train,
  cluster.data,
  cluster.markers,
  scale = TRUE
)

Arguments

Details

The function operates as follows: 1. Ensures that the 'cluster.data' has the correct column names. 2. Adds a one-hot encoded matrix for cluster assignments. 3. Scales the test data for prediction if 'scale' is TRUE. 4. Selects genes that are common between the test and training datasets. 5. Uses glmnet to identify the important markers for each cluster and trains a Random Forest model for classification. 6. Predicts the cluster for test samples and provides probabilities for each cluster.

Value

A data frame with: - ID: The sample identifier. - Cluster: The predicted cluster label for each sample. - Probabilities: The probabilities for each cluster assignment.

Author(s)

Zaoqu Liu; Email: [email protected]

Examples

cluster.data <- data.frame(
  Sample = paste0("Sample", 1:60),
  Cluster = rep(paste0("C", 1:3), each = 20)
)
data.train <- matrix(rnorm(6000),
  nrow = 100,
  dimnames = list(paste0("Gene", 1:100), cluster.data$Sample)
)
data.test <- matrix(rnorm(5000),
  nrow = 100,
  dimnames = list(paste0("Gene", 1:100), paste0("P", 1:50))
)
cluster.markers <- setNames(
  lapply(
    unique(cluster.data$Cluster),
    function(cluster) {
      data.frame(Gene = sample(rownames(data.train), 10))
    }
  ),
  unique(cluster.data$Cluster)
)
result <- Classifier.RF(
  data.test = data.test, data.train = data.train,
  cluster.data, cluster.markers
)
head(result)

Description

This function performs classification using Ridge Regression to predict cluster assignments for test data based on trained models from training data and cluster markers.

Usage

Classifier.Ridge(
  data.test,
  data.train,
  cluster.data,
  cluster.markers,
  scale = TRUE
)

Arguments

Details

The function operates as follows: 1. Ensures that the 'cluster.data' has the correct column names. 2. Adds a one-hot encoded matrix for cluster assignments. 3. Scales the test data for prediction. 4. Selects genes that are common between the test and training datasets. 5. Uses glmnet to identify the important markers for each cluster and trains a Ridge model for classification. 6. Predicts the cluster for test samples and provides probabilities for each cluster.

Value

A data frame with: - ID: The sample identifier. - Probabilities: The probabilities for each cluster assignment. - Predict: The predicted cluster label for each sample.

Author(s)

Zaoqu Liu; Email: [email protected]

Examples

cluster.data <- data.frame(
  Sample = paste0("Sample", 1:60),
  Cluster = rep(paste0("C", 1:3), each = 20)
)
data.train <- matrix(rnorm(6000),
  nrow = 100,
  dimnames = list(paste0("Gene", 1:100), cluster.data$Sample)
)
data.test <- matrix(rnorm(5000),
  nrow = 100,
  dimnames = list(paste0("Gene", 1:100), paste0("P", 1:50))
)
cluster.markers <- setNames(
  lapply(
    unique(cluster.data$Cluster),
    function(cluster) {
      data.frame(Gene = sample(rownames(data.train), 10))
    }
  ),
  unique(cluster.data$Cluster)
)
result <- Classifier.Ridge(
  data.test = data.test, data.train = data.train,
  cluster.data, cluster.markers
)
head(result)

Description

This function performs single-sample Gene Set Enrichment Analysis (ssGSEA) to assign subtypes to samples based on marker gene sets. It calculates enrichment scores with permutation testing and predicts sample subtypes based on the most significant enrichment results.

Usage

Classifier.ssGSEA(
  data.test,
  marker.list,
  dir.file = ".",
  gct.filename = "data.gct",
  number.perms = 100,
  tolerate.mixed = FALSE,
  method = c("internal", "GSVA", "external"),
  seed = 12345
)

Arguments

Details

The function:

1. Calculates ssGSEA enrichment scores for each marker gene set in every sample.

2. Performs permutation testing to estimate statistical significance.

3. Predicts subtypes by identifying the marker gene set with the most significant enrichment (smallest p-value).

4. If tolerate.mixed is TRUE and multiple gene sets share the same minimum p-value, the sample is labeled as "Mixed".

Value

A data frame with the following columns:

• ID: Sample identifiers.

• Predict: Predicted subtype for each sample.

• Columns with _pval: P-values for each marker gene set or subtype.

Author(s)

Zaoqu Liu; Email: [email protected]

References

Wang Q, Hu B, Hu X, Kim H, Squatrito M, Scarpace L, et al. Tumor Evolution of Glioma-Intrinsic Gene Expression Subtypes Associates with Immunological Changes in the Microenvironment. Cancer Cell. July 2017;32(1):42-56.e6.

Examples

# Simulated expression data
data.test <- matrix(rnorm(10000), nrow = 100, ncol = 100)
rownames(data.test) <- paste0("Gene", 1:100)
colnames(data.test) <- paste0("Sample", 1:100)

# Example marker list
marker.list <- list(
  Subtype1 = c("Gene1", "Gene2", "Gene3"),
  Subtype2 = c("Gene4", "Gene5", "Gene6")
)

# Run ssGSEA-based subtyping
result <- Classifier.ssGSEA(
  data.test = data.test,
  marker.list = marker.list,
  number.perms = 50,
  tolerate.mixed = TRUE
)
print(result)

Description

This function performs classification using Stepwise Logistic Regression to predict cluster assignments for test data based on trained models from training data and cluster markers.

Usage

Classifier.StepLR(
  data.test,
  data.train,
  cluster.data,
  cluster.markers,
  scale = TRUE
)

Arguments

Details

The function operates as follows: 1. Ensures that the 'cluster.data' has the correct column names. 2. Scales the test data for prediction. 3. Selects genes that are common between the test and training datasets. 4. Uses glmnet to identify the important markers for each cluster and trains a multinomial logistic regression model for classification. 5. Predicts the cluster for test samples and provides probabilities for each cluster.

Value

A data frame with: - ID: The sample identifier. - Probabilities: The probabilities for each cluster assignment. - Predict: The predicted cluster label for each sample.

Author(s)

Zaoqu Liu; Email: [email protected]

Examples

cluster.data <- data.frame(
  Sample = paste0("Sample", 1:60),
  Cluster = rep(paste0("C", 1:3), each = 20)
)
data.train <- matrix(rnorm(6000),
  nrow = 100,
  dimnames = list(paste0("Gene", 1:100), cluster.data$Sample)
)
data.test <- matrix(rnorm(5000),
  nrow = 100,
  dimnames = list(paste0("Gene", 1:100), paste0("P", 1:50))
)
cluster.markers <- setNames(
  lapply(
    unique(cluster.data$Cluster),
    function(cluster) {
      data.frame(Gene = sample(rownames(data.train), 10))
    }
  ),
  unique(cluster.data$Cluster)
)
result <- Classifier.StepLR(
  data.test = data.test, data.train = data.train,
  cluster.data, cluster.markers
)
head(result)

Description

This function performs classification using SVD and a neural network to predict cluster assignments for test data based on trained models from training data and cluster markers.

Usage

Classifier.SVD(
  data.test,
  data.train,
  cluster.data,
  cluster.markers,
  scale = TRUE
)

Arguments

Details

The function operates as follows: 1. Ensures that the 'cluster.data' has the correct column names. 2. Adds a one-hot encoded matrix for cluster assignments. 3. Scales the test data for prediction. 4. Selects genes that are common between the test and training datasets. 5. Uses glmnet to identify the important markers for each cluster and performs SVD to reduce dimensionality. 6. Trains a neural network model for classification using the top singular vectors. 7. Predicts the cluster for test samples and provides probabilities for each cluster.

Value

A data frame with: - ID: The sample identifier. - Probabilities: The probabilities for each cluster assignment. - Predict: The predicted cluster label for each sample.

Author(s)

Zaoqu Liu; Email: [email protected]

Examples

cluster.data <- data.frame(
  Sample = paste0("Sample", 1:60),
  Cluster = rep(paste0("C", 1:3), each = 20)
)
data.train <- matrix(rnorm(6000),
  nrow = 100,
  dimnames = list(paste0("Gene", 1:100), cluster.data$Sample)
)
data.test <- matrix(rnorm(5000),
  nrow = 100,
  dimnames = list(paste0("Gene", 1:100), paste0("P", 1:50))
)
cluster.markers <- setNames(
  lapply(
    unique(cluster.data$Cluster),
    function(cluster) {
      data.frame(Gene = sample(rownames(data.train), 10))
    }
  ),
  unique(cluster.data$Cluster)
)
result <- Classifier.SVD(
  data.test = data.test, data.train = data.train,
  cluster.data, cluster.markers
)
head(result)

Description

This function performs classification using Support Vector Machine (SVM) to predict cluster assignments for test data based on trained models from training data and cluster markers.

Usage

Classifier.SVM(
  data.test,
  data.train,
  cluster.data,
  cluster.markers,
  scale = TRUE
)

Arguments

Details

The function operates as follows: 1. Ensures that the 'cluster.data' has the correct column names. 2. Adds a one-hot encoded matrix for cluster assignments. 3. Scales the test data for prediction. 4. Selects genes that are common between the test and training datasets. 5. Uses glmnet to identify the important markers for each cluster and trains an SVM model for classification. 6. Predicts the cluster for test samples and provides probabilities for each cluster.

Value

A data frame with: - ID: The sample identifier. - Probabilities: The probabilities for each cluster assignment. - Predict: The predicted cluster label for each sample.

Author(s)

Zaoqu Liu; Email: [email protected]

Examples

cluster.data <- data.frame(
  Sample = paste0("Sample", 1:60),
  Cluster = rep(paste0("C", 1:3), each = 20)
)
data.train <- matrix(rnorm(6000),
  nrow = 100,
  dimnames = list(paste0("Gene", 1:100), cluster.data$Sample)
)
data.test <- matrix(rnorm(5000),
  nrow = 100,
  dimnames = list(paste0("Gene", 1:100), paste0("P", 1:50))
)
cluster.markers <- setNames(
  lapply(
    unique(cluster.data$Cluster),
    function(cluster) {
      data.frame(Gene = sample(rownames(data.train), 10))
    }
  ),
  unique(cluster.data$Cluster)
)
result <- Classifier.SVM(
  data.test = data.test, data.train = data.train,
  cluster.data, cluster.markers
)
head(result)

Description

This function performs classification using XGBoost to predict cluster assignments for test data based on trained models from training data and cluster markers.

Usage

Classifier.XGBoost(
  data.test,
  data.train,
  cluster.data,
  cluster.markers,
  scale = TRUE
)

Arguments

Details

The function operates as follows: 1. Ensures that the 'cluster.data' has the correct column names. 2. Adds a one-hot encoded matrix for cluster assignments. 3. Scales the test data for prediction. 4. Selects genes that are common between the test and training datasets. 5. Uses glmnet to identify the important markers for each cluster and trains an XGBoost model for classification. 6. Predicts the cluster for test samples and provides probabilities for each cluster.

Value

A data frame with: - ID: The sample identifier. - Probabilities: The probabilities for each cluster assignment. - Predict: The predicted cluster label for each sample.

Author(s)

Zaoqu Liu; Email: [email protected]

Examples

cluster.data <- data.frame(
  Sample = paste0("Sample", 1:60),
  Cluster = rep(paste0("C", 1:3), each = 20)
)
data.train <- matrix(rnorm(6000),
  nrow = 100,
  dimnames = list(paste0("Gene", 1:100), cluster.data$Sample)
)
data.test <- matrix(rnorm(5000),
  nrow = 100,
  dimnames = list(paste0("Gene", 1:100), paste0("P", 1:50))
)
cluster.markers <- setNames(
  lapply(
    unique(cluster.data$Cluster),
    function(cluster) {
      data.frame(Gene = sample(rownames(data.train), 10))
    }
  ),
  unique(cluster.data$Cluster)
)
result <- Classifier.XGBoost(
  data.test = data.test, data.train = data.train,
  cluster.data, cluster.markers
)
head(result)

Description

Creates a comparison matrix showing agreement between algorithms.

Usage

compare_clusterings(results)

Arguments

Value

Matrix of Adjusted Rand Index values.

Description

Performs UMAP for visualization of multi-omics data.

Usage

compute_umap(data, n_neighbors = 15, min_dist = 0.1, n_epochs = 200, seed = 42)

Arguments

Value

Matrix with UMAP coordinates (samples x 2).

Description

Performs consensus clustering to identify stable clusters.

Usage

consensus_cluster(
  d,
  maxK = 6,
  reps = 1000,
  pItem = 0.8,
  pFeature = 1,
  clusterAlg = "hc",
  innerLinkage = "ward.D2",
  finalLinkage = "ward.D2",
  distance = "euclidean",
  seed = NULL,
  verbose = FALSE
)

Arguments

Value

List containing consensus matrices and clustering results.

Description

Performs simple batch correction using mean centering.

Usage

correct_batch(data_list, batch, method = "center")

Arguments

Value

List of batch-corrected matrices.

Description

Performs Consensus PCA on multiple data types.

Usage

cpca(data_list, ncomp = 2)

Arguments

Value

List with scores, loadings, eigenvalues.

Description

This function calculates the coefficient of variation (CV) for a given numeric vector.

Usage

CV(V)

Arguments

Value

The coefficient of variation (CV) for the input vector.

Author(s)

Zaoqu Liu; E-mail: [email protected]

Description

This function calculates the coefficient of variation (CV) for each column in a data frame.

Usage

CV.df(df)

Arguments

Value

A sorted vector of CV values for each column in the data frame, in decreasing order.

Author(s)

Zaoqu Liu; E-mail: [email protected]

Description

This function performs feature selection using a logistic regression model and bootstrapping for each cluster in the dataset. For each cluster, the function evaluates which features are significantly associated with the cluster based on a logistic regression model with bootstrap sampling. The output includes a count of significant features for each cluster based on the p-value threshold.

Usage

FeatureSelectionWithBootstrap(
  data,
  p.no_bootstrap = 0.01,
  p.bootstrap = 0.05,
  num.iteration = 1000,
  nCores = parallel::detectCores() - 3
)

Arguments

Value

A list of data frames, each containing the features selected for each cluster and the count of significant features based on bootstrapping.

Author(s)

Zaoqu Liu; Email: [email protected]

Description

Keeps features with highest MAD values.

Usage

filter_by_mad(data_list, top_n = 5000)

Arguments

Value

List of filtered matrices.

Description

Removes features with low variance across samples.

Usage

filter_low_variance(data_list, min_var = 0.01, top_n = NULL, top_pct = NULL)

Arguments

Value

List of filtered matrices.

Description

Selects optimal feature combinations for multi-modality clustering analysis.

Usage

Find.OptClusterFeatures(
  data_layers,
  feature_subset_sizes,
  try_num_clusters = 2:6,
  n_runs = 5,
  n_fold = 5
)

Arguments

Value

A list with optimal_combination and all_results.

Author(s)

Zaoqu Liu; Email: [email protected]

Description

The gene_sets data object contains a unified collection of gene sets derived from multiple sources, including curated pathways (c2.cp.v2022.1.Hs), Gene Ontology terms (c5.go.v2022.1.Hs), and hallmark gene sets (h.all.v2022.1.Hs). These gene sets are based on version 2022.1 and represent key biological processes, pathways, and well-defined biological states. By integrating these sources, the gene_sets object provides a comprehensive dataset that can be utilized for enrichment analysis and functional exploration, offering valuable insights into underlying biological mechanisms.

Usage

gene_sets

Format

data.frame

Description

Extracts cluster assignments for a specific K.

Usage

get_consensus_class(consensus_fit, k)

Arguments

Value

Data frame with sample IDs and cluster assignments.

Description

This function extracts binary cluster assignments from multiple clustering results.

Usage

get.binary.clusters(res)

Arguments

Value

A data frame where each row represents a binary encoding of cluster assignments across different methods.

Author(s)

Zaoqu Liu; Email: [email protected]

Description

Extract cluster assignments from Consensus Clustering results for a specific number of clusters.

Usage

get.class(cc.res, k)

Arguments

Value

A data frame with the following columns: - ID: The sample identifier. - Cluster: The assigned cluster label, prefixed by 'C'.

Author(s)

Zaoqu Liu; Email: [email protected]

Examples

data <- mtcars
cc_res <- RunCC(data)
clu <- get.class(cc_res, 2)
clu

Description

This function calculates the Jaccard distance or similarity for a binary matrix. It is typically used to evaluate the similarity or dissimilarity between columns of a binary matrix.

Usage

get.Jaccard.Distance(data, dissimilarity = TRUE)

Arguments

Details

The function computes the Jaccard distance (or similarity) between each pair of columns in the input binary matrix. The Jaccard distance is calculated as 1 minus the Jaccard similarity.

Value

A matrix containing the Jaccard distance or similarity between each pair of columns in the input matrix.

Author(s)

Zaoqu Liu; Email: [email protected]

Examples

data <- matrix(sample(0:1, 1500, replace = TRUE), nrow = 30, ncol = 50)
jaccard_dist <- get.Jaccard.Distance(as.data.frame(data), dissimilarity = TRUE)
jaccard_dist

Description

Imputes or removes missing values in multi-omics data.

Usage

handle_missing(data_list, method = "median", threshold = 0.5, k = 5)

Arguments

Value

List of processed matrices.

Description

Faster version using variational approximation.

Usage

icluster_bayes_fast(data_list, k, max_iter = 100, tol = 1e-04, verbose = FALSE)

Arguments

Value

List with cluster assignments and parameters.

Description

Installs optional packages for extended functionality.

Usage

init(classifier = TRUE, survival = FALSE, gsva = FALSE)

Arguments

Author(s)

Zaoqu Liu; Email: [email protected]

Description

Main IntNMF algorithm using non-negative alternating least squares.

Usage

intnmf_cluster(
  dat,
  k,
  maxiter = 200,
  st_count = 20,
  n_ini = 30,
  ini_nndsvd = TRUE,
  seed = TRUE,
  wt = NULL
)

Arguments

Value

List with W, H matrices, clusters, and convergence info.

Description

Selects optimal number of clusters using cross-validation.

Usage

intnmf_opt_k(
  dat,
  n_runs = 30,
  n_fold = 5,
  k_range = 2:8,
  maxiter = 100,
  st_count = 10,
  wt = NULL,
  verbose = TRUE
)

Arguments

Value

Matrix of CPI values for each K and run.

Description

Performs low-rank approximation clustering.

Usage

lracluster(data, types, dimension = 2, names = NULL)

Arguments

Value

List with coordinate matrix and potential (quality measure).

Description

This function calculates the median absolute deviation (MAD) for each column in a data frame.

Usage

MAD.df(df)

Arguments

Value

A sorted vector of MAD values for each column in the data frame, in decreasing order.

Author(s)

Zaoqu Liu; E-mail: [email protected]

Description

Performs MCIA on multiple data matrices.

Usage

mcia(data_list, ncomp = 2)

Arguments

Value

List with global scores, block scores, loadings, and eigenvalues.

Description

This function calculates the mean value for each column in a data frame.

Usage

Mean.df(df)

Arguments

Value

A vector of mean values for each column in the data frame.

Author(s)

Zaoqu Liu; E-mail: [email protected]

Description

This function calculates the median value for each column in a data frame.

Usage

Median.df(df)

Arguments

Value

A vector of median values for each column in the data frame.

Author(s)

Zaoqu Liu; E-mail: [email protected]

Description

This function performs min-max normalization on a numeric vector.

Usage

minmax(x)

Arguments

Value

A normalized numeric vector with values between 0 and 1.

Author(s)

Zaoqu Liu; E-mail: [email protected]

Description

This function performs min-max normalization on each column in a data frame.

Usage

minmax.df(data)

Arguments

Value

A data frame with normalized values between 0 and 1 for each column.

Author(s)

Zaoqu Liu; E-mail: [email protected]

Description

Constructs a single affinity graph from multi-omics data.

Usage

nemo_affinity_graph(raw_data, k = NA)

Arguments

Value

Affinity matrix measuring similarity across all omics.

Description

Performs multi-omic clustering using the NEMO algorithm.

Usage

nemo_clustering(omics_list, num_clusters = NULL, num_neighbors = NA)

Arguments

Value

Named vector of cluster assignments.

Description

Uses eigengap heuristic to estimate optimal cluster number.

Usage

nemo_num_clusters(W, NUMC = 2:15)

Arguments

Value

Estimated number of clusters.

Description

Functions for parallel execution of multi-omics analysis.

Sets up parallel processing using the future framework.

Usage

setup_parallel(workers = NULL, strategy = "multisession", verbose = TRUE)

Arguments

Details

- "multisession": Works on all platforms (Windows, macOS, Linux) - "multicore": Faster on Unix-like systems, not available on Windows - "sequential": No parallelism (useful for debugging)

Value

Invisibly returns the previous plan.

Author(s)

Zaoqu Liu; Email: [email protected]

Description

Performs bootstrap-based feature selection in parallel.

Usage

parallel_bootstrap_features(
  data,
  n_bootstrap = 1000,
  n_workers = NULL,
  p_threshold = 0.05,
  progress = TRUE
)

Arguments

Value

List of significant features per cluster.

Description

Performs consensus clustering with parallel resampling.

Usage

parallel_consensus_cluster(
  data,
  max_k = 6,
  n_reps = 1000,
  n_workers = NULL,
  prop_sample = 0.8,
  cluster_method = "hclust",
  verbose = TRUE
)

Arguments

Value

List with consensus matrices and cluster results.

Description

This function performs pathway differential expression analysis based on clustering results and pathway activity scores derived from ssMwwGST.

Usage

PathDEA(
  Cluster_data,
  ssMwwGST_results,
  dea_FDR_threshold = 0.001,
  dea_gap_threshold = 1.5
)

Arguments

Details

The function operates as follows: 1. Extracts Normalized Enrichment Scores (NES) from the ssMwwGST results. 2. Performs Wilcoxon rank-sum tests to compare pathway activity between clusters for each pathway. 3. Calculates median and mean differences in pathway activity between clusters. 4. Adjusts p-values using the Benjamini-Hochberg method to control the false discovery rate (FDR).

Value

A list containing: - dea_path: A list of data frames, each containing the differential expression analysis results for each cluster. - dea_path2: A list of data frames containing the filtered differential expression analysis results for each cluster based on FDR and median gap thresholds. - NES: A data frame of Normalized Enrichment Scores for each gene set and each sample. - Cluster: A data frame of sample IDs and their corresponding cluster assignments.

Author(s)

Zaoqu Liu; Email: [email protected]

Examples

# Example usage:
Cluster_data <- data.frame(Sample = paste0("Sample", 1:10), Cluster = rep(1:2, each = 5))
ssMwwGST_results <- list(NES = matrix(rnorm(100),
  nrow = 10, ncol = 10,
  dimnames = list(paste0("Pathway", 1:10), paste0("Sample", 1:10))
))
result <- PathDEA(Cluster_data, ssMwwGST_results)

Description

Performs perturbation-based clustering to find optimal k.

Usage

perturbation_clustering(
  data,
  kMin = 2,
  kMax = 5,
  k = NULL,
  n_iter = 50,
  clustering_method = "kmeans",
  verbose = FALSE
)

Arguments

Value

List with k, cluster, origS, pertS.

Description

Visualizes agreement between different clustering algorithms.

Usage

plot_algorithm_comparison(results, title = "Algorithm Agreement (ARI)")

Arguments

Value

A ggplot object or base R plot.

Description

Visualizes PAC, CHI, and silhouette scores across different K values.

Usage

plot_cluster_quality(
  pac_values,
  chi_values = NULL,
  title = "Cluster Quality Assessment"
)

Arguments

Value

A ggplot object or base R plot.

Description

Visualizes the consensus matrix from consensus clustering.

Usage

plot_consensus_heatmap(
  consensus_matrix,
  clusters = NULL,
  title = "Consensus Matrix",
  colors = NULL
)

Arguments

Value

Invisibly returns the plot object.

Description

Creates a silhouette plot for cluster quality evaluation.

Usage

plot_silhouette(
  clusters,
  dist_matrix,
  title = "Silhouette Analysis",
  colors = NULL
)

Arguments

Value

A ggplot object or base R plot.

Description

Visualizes survival differences between clusters.

Usage

plot_survival(
  time,
  event,
  clusters,
  title = "Kaplan-Meier Survival Curves",
  colors = NULL,
  conf_int = TRUE,
  risk_table = FALSE
)

Arguments

Value

A ggplot object (if survminer available) or base R plot.

Description

Creates a UMAP plot colored by cluster assignments.

Usage

plot_umap(
  umap_coords,
  clusters,
  title = "UMAP Visualization",
  point_size = 2,
  colors = NULL,
  show_legend = TRUE
)

Arguments

Value

A ggplot object if ggplot2 is available, otherwise base R plot.

Description

Functions for preprocessing multi-omics data before integration.

Applies normalization to each data matrix in the list.

Usage

normalize_omics(data_list, method = "zscore", by_feature = TRUE)

Arguments

Value

List of normalized matrices.

Author(s)

Zaoqu Liu

Description

Generates quality control summary for multi-omics data.

Usage

qc_summary(data_list)

Arguments

Value

Data frame with QC metrics for each dataset.

Description

Performs RGCCA for multi-block data analysis.

Usage

rgcca(
  A,
  C = 1 - diag(length(A)),
  tau = rep(1, length(A)),
  ncomp = rep(1, length(A)),
  scheme = "centroid",
  scale = TRUE,
  init = "svd",
  bias = TRUE,
  tol = 1e-08,
  verbose = FALSE
)

Arguments

Value

List with Y, a, astar, C, tau, scheme, ncomp, crit, AVE.

Description

Performs BCC-based clustering on multi-omics data.

Usage

run_bcc(data, n_clusters, method = "fast", n_iter = 1000)

Arguments

Value

Data frame with sample IDs and cluster assignments.

Description

Performs CIMLR-based clustering on multi-omics data.

Usage

run_cimlr(data, n_clusters, n_kernels = 10)

Arguments

Value

Data frame with sample IDs and cluster assignments.

Description

Performs CPCA-based clustering on multi-omics data.

Usage

run_cpca(data, n_clusters, ncomp = NULL, cluster_method = "ward.D2")

Arguments

Value

Data frame with cluster assignments.

Description

Performs iClusterBayes-based clustering on multi-omics data.

Usage

run_iclusterbayes(
  data,
  n_clusters,
  method = "fast",
  n_burnin = 500,
  n_sample = 1000
)

Arguments

Value

Data frame with sample IDs and cluster assignments.

Description

Unified function to run any multi-omics clustering algorithm.

Usage

run_integration(data, algorithm, n_clusters, ...)

Arguments

Value

Data frame with sample IDs, cluster assignments, and cluster labels.

Examples

## Not run: 
# Create example data
data1 <- matrix(rnorm(100 * 50), nrow = 100, ncol = 50)
data2 <- matrix(rnorm(80 * 50), nrow = 80, ncol = 50)
colnames(data1) <- colnames(data2) <- paste0("Sample", 1:50)

data_list <- list(GE = data1, ME = data2)

# Run SNF clustering
result <- run_integration(data_list, "SNF", n_clusters = 3)

## End(Not run)

Description

High-level function for IntNMF clustering on multi-omics data.

Usage

run_intnmf(data, n_clusters, maxiter = 200, n_init = 30, weight = NULL)

Arguments

Value

Data frame with sample IDs and cluster assignments.

Description

Ensemble clustering by combining single-omics results.

Usage

run_late_fusion(
  data,
  n_clusters,
  single_method = "kmeans",
  consensus_method = "similarity"
)

Arguments

Value

Data frame with cluster assignments.

Description

Performs LRAcluster-based clustering on multi-omics data.

Usage

run_lracluster(data, n_clusters, types = NULL, cluster_method = "ward.D2")

Arguments

Value

Data frame with sample IDs and cluster assignments.

Description

Performs MCIA-based clustering on multi-omics data.

Usage

run_mcia(data, n_clusters, ncomp = NULL, cluster_method = "ward.D2")

Arguments

Value

Data frame with sample IDs and cluster assignments.

Description

Performs factor analysis based clustering on multi-omics data. Uses multi-view factor analysis to extract latent factors, then clusters samples based on these factors. For full MOFA, use MOFA2 package.

Usage

run_mofa(data, n_clusters, n_factors = 10, cluster_method = "ward.D2")

Arguments

Value

Data frame with sample IDs and cluster assignments.

Description

Runs multiple algorithms and returns combined results.

Usage

run_multiple_algorithms(data, algorithms = NULL, n_clusters, ...)

Arguments

Value

List of results for each algorithm.

Description

Performs NEMO-based clustering on multi-omics data.

Usage

run_nemo(data, n_clusters = NULL, n_neighbors = NA)

Arguments

Value

Data frame with sample IDs and cluster assignments.

Description

Executes multiple clustering algorithms in parallel.

Usage

run_parallel_algorithms(
  data,
  algorithms = NULL,
  n_clusters,
  n_workers = NULL,
  ...
)

Arguments

Value

List of results for each algorithm.

Description

Performs PINSPlus-based clustering on multi-omics data.

Usage

run_pinsplus(data, n_clusters = NULL, kMin = 2, kMax = 5)

Arguments

Value

Data frame with sample IDs and cluster assignments.

Description

Performs RGCCA-based clustering on multi-omics data.

Usage

run_rgcca(
  data,
  n_clusters,
  ncomp = NULL,
  tau = NULL,
  scheme = "centroid",
  cluster_method = "ward.D2"
)

Arguments

Value

Data frame with sample IDs and cluster assignments.

Description

Performs SGCCA-based clustering on multi-omics data.

Usage

run_sgcca(
  data,
  n_clusters,
  ncomp = NULL,
  c1 = NULL,
  scheme = "centroid",
  cluster_method = "ward.D2"
)

Arguments

Value

Data frame with cluster assignments.

Description

Performs SNF-based clustering on multi-omics data.

Usage

run_snf(data, n_clusters, n_neighbors = 20, sigma = 0.5, n_iterations = 20)

Arguments

Value

Data frame with sample IDs, cluster assignments, and cluster labels.

Description

SNF with custom weights for each data type.

Usage

run_wsnf(
  data,
  n_clusters,
  weights = NULL,
  n_neighbors = 20,
  sigma = 0.5,
  n_iterations = 20
)

Arguments

Value

Data frame with cluster assignments.

Description

Provides a unified interface for all clustering algorithms in MOFSR.

Returns available multi-omics clustering algorithms.

Usage

list_clustering_algorithms()

Value

Character vector of algorithm names.

Author(s)

Zaoqu Liu

Description

Performs multi-omics clustering using BCC algorithm.

Usage

RunBCC(data, N.clust, method = "fast", max.iterations = 1000)

Arguments

Value

A data frame with Sample, Cluster, and Cluster2 columns.

Author(s)

Zaoqu Liu; Email: [email protected]

References

Lock EF, Dunson DB. Bioinformatics. 2013.

Description

Performs consensus clustering on the given data.

Usage

RunCC(
  data,
  maxK = 6,
  reps = 1000,
  pItem = 0.8,
  pFeature = 1,
  clusterAlg = "hc",
  distance = "euclidean",
  innerLinkage = "ward.D2",
  finalLinkage = "ward.D2",
  seed = 1234,
  verbose = FALSE
)

Arguments

Value

A list containing consensus clustering results.

Author(s)

Zaoqu Liu; Email: [email protected]

Description

Performs multi-omics clustering using CIMLR algorithm.

Usage

RunCIMLR(data, N.clust, n.kernels = 10)

Arguments

Value

A data frame with Sample, Cluster, and Cluster2 columns.

Author(s)

Zaoqu Liu; Email: [email protected]

References

Ramazzotti D, et al. Nature Communications. 2018.

Description

This function runs different classification models based on user input to predict cluster assignments for test data.

Usage

RunClassifier(
  algorithm,
  data.test,
  data.train,
  cluster.data,
  cluster.markers,
  scale = TRUE
)

Arguments

Details

The function dynamically selects and runs a classification model based on user input. The supported classifiers include a range of machine learning models such as Random Forest, kNN, PCA, SVM, LASSO, Ridge, and others.

Value

A data frame containing the prediction results based on the selected algorithm.

Author(s)

Zaoqu Liu; Email: [email protected]

Examples

# Example usage:
data.test <- matrix(rnorm(1000), nrow = 100, ncol = 10)
data.train <- matrix(rnorm(1000), nrow = 100, ncol = 10)
cluster.data <- data.frame(Sample = paste0("Sample", 1:10), Cluster = rep(1:2, each = 5))
cluster.markers <- setNames(lapply(unique(cluster.data$Cluster), function(c) data.frame(Gene = paste0("Gene", sample(1:30, 10)), OR = runif(10, 0.5, 2), AUC = runif(10))), unique(cluster.data$Cluster))
result <- RunClassifier(algorithm = "RF", data.test, data.train, cluster.data, cluster.markers, scale = TRUE)

Description

Performs Consensus Clustering Analysis to identify stable clusters.

Usage

RunCOCA(
  jaccard.matrix,
  max.clusters = 6,
  optimal.clusters = 3,
  linkage.method = "ward.D2",
  clustering.algorithm = "hc",
  distance.metric = "euclidean",
  resampling.iterations = 10000,
  resample.proportion = 0.7
)

Arguments

Value

A list with consensus clustering results.

Author(s)

Zaoqu Liu; Email: [email protected]

Description

Performs multi-omics clustering using CPCA algorithm.

Usage

RunCPCA(data, N.clust, ncomp = NULL, clustering.algorithm = "ward.D2")

Arguments

Value

A data frame with Sample, Cluster, and Cluster2 columns.

Author(s)

Zaoqu Liu; Email: [email protected]

Description

This function runs an ensemble of different classification models to predict cluster assignments for test data, considering consensus among the models.

Usage

RunEnsemble(
  data.test,
  data.train,
  cluster.data,
  cluster.markers,
  surdata = NULL,
  time = "time",
  event = "event",
  methods = NULL,
  sur.trend.rank = NULL,
  cutoff.P = 0.05
)

Arguments

Details

This function runs an ensemble of classifiers, checks the consistency among classifiers, and optionally performs survival analysis to filter models based on trends in clinical outcomes.

Value

A list containing the ensemble prediction results and optional survival analysis.

Author(s)

Zaoqu Liu; Email: [email protected]

Examples

# Example usage:
data.test <- matrix(rnorm(1000), nrow = 100, ncol = 10)
data.train <- matrix(rnorm(1000), nrow = 100, ncol = 10)
cluster.data <- data.frame(Sample = paste0("Sample", 1:10), Cluster = rep(1:2, each = 5))
cluster.markers <- setNames(lapply(unique(cluster.data$Cluster), function(c) data.frame(Gene = paste0("Gene", sample(1:30, 10)), OR = runif(10, 0.5, 2), AUC = runif(10))), unique(cluster.data$Cluster))
surdata <- data.frame(ID = paste0("Sample", 1:10), time = runif(10, 1, 1000), event = sample(0:1, 10, replace = TRUE))
result <- RunEnsemble(data.test, data.train, cluster.data, cluster.markers, surdata, time = "time", event = "event", methods = c("rf", "xgboost"), sur.trend.rank = c("C1", "C3", "C2"))

Description

This function estimates gene-set enrichment scores across all samples using various methods.

Usage

RunGSVA(
  exp,
  gene.list,
  min.size = 3,
  max.size = 1000,
  method = "ssgsea",
  ssgsea.normalize = TRUE,
  ssgsea.alpha = 0.25,
  gsva.kcdf = "Gaussian",
  gsva.tau = 1,
  gsva.maxDiff = TRUE,
  gsva.absRanking = FALSE,
  verbose = TRUE,
  nCores = parallel::detectCores() - 3
)

Arguments

Details

This function supports multiple methods for estimating gene-set enrichment scores, including ssGSEA, GSVA, zscore, and plage. The scores are calculated for each gene set across all samples. The 'ssGSES' function is flexible and allows for customization of the minimum and maximum size of gene sets considered in the analysis. By providing different methods, the function can adapt to various types of gene-set enrichment analysis, each having its own strengths and suitable applications.

• "gsva": Gene Set Variation Analysis, suitable for detecting subtle changes in pathway activity.

• "ssgsea": Single-Sample Gene Set Enrichment Analysis, useful for individual sample analysis.

• "zscore": Z-score transformation, a simpler approach to standardize expression values.

• "plage": Pathway Level Analysis of Gene Expression, which focuses on correlating pathway components.

Value

A gene-set by sample matrix of gene-set enrichment scores.

Author(s)

Zaoqu Liu; Email: [email protected]

Description

Performs multi-omics clustering using iClusterBayes algorithm.

Usage

RuniClusterBayes(data, N.clust, method = "fast", burnin = 500, nsample = 1000)