@bgicli/bgicli 2.1.1 → 2.2.0

package/data/skills/bio-multi-omics-mixomics-analysis/SKILL.md

@@ -0,0 +1,221 @@
+---
+name: bio-multi-omics-mixomics-analysis
+description: Supervised and unsupervised multi-omics integration with mixOmics. Includes sPLS for pairwise integration and DIABLO for multi-block discriminant analysis. Use when performing supervised multi-omics integration or identifying features that discriminate between groups.
+tool_type: r
+primary_tool: mixOmics
+---
+
+## Version Compatibility
+
+Reference examples tested with: mixOmics 6.26+
+
+Before using code patterns, verify installed versions match. If versions differ:
+- R: `packageVersion("<pkg>")` then `?function_name` to verify parameters
+
+If code throws ImportError, AttributeError, or TypeError, introspect the installed
+package and adapt the example to match the actual API rather than retrying.
+
+# mixOmics Multi-Omics Analysis
+
+**"Integrate my multi-omics data with supervised analysis"** → Identify cross-omics feature signatures that discriminate between groups using sparse PLS and multi-block discriminant analysis.
+- R: `mixOmics::block.splsda()` (DIABLO), `mixOmics::spls()` for pairwise integration
+
+## Setup and Data Preparation
+
+**Goal:** Load and align omics matrices with matching sample labels and phenotype information.
+
+**Approach:** Read each omics layer and phenotype, then intersect to common samples.
+
+```r
+library(mixOmics)
+
+# Load omics matrices (samples x features)
+X_rna <- as.matrix(read.csv('rnaseq.csv', row.names = 1))
+X_protein <- as.matrix(read.csv('proteomics.csv', row.names = 1))
+Y <- factor(read.csv('phenotype.csv')$Condition)
+
+# Ensure matching samples
+common <- Reduce(intersect, list(rownames(X_rna), rownames(X_protein)))
+X_rna <- X_rna[common, ]
+X_protein <- X_protein[common, ]
+Y <- Y[match(common, read.csv('phenotype.csv')$Sample)]
+```
+
+## Pairwise Integration: sPLS
+
+**Goal:** Identify correlated features between two omics layers using sparse partial least squares.
+
+**Approach:** Tune component count, fit sPLS with feature selection (keepX/keepY), and visualize cross-omics correlations.
+
+```r
+# Sparse Partial Least Squares for two datasets
+# Finds correlated features between omics
+
+# Tune number of components
+tune_spls <- perf(spls(X_rna, X_protein, ncomp = 5), validation = 'Mfold', folds = 5)
+plot(tune_spls)
+
+# Run sPLS
+spls_result <- spls(X_rna, X_protein, ncomp = 3, keepX = c(50, 50, 50), keepY = c(30, 30, 30))
+
+# Visualize correlations
+plotIndiv(spls_result, comp = c(1, 2), group = Y, legend = TRUE)
+plotVar(spls_result, comp = c(1, 2), var.names = TRUE)
+
+# Correlation circle
+plotArrow(spls_result, group = Y)
+
+# Heatmap of selected features
+cim(spls_result, comp = 1)
+```
+
+## DIABLO: Multi-Block Discriminant Analysis
+
+**Goal:** Find multi-omics feature signatures that discriminate between experimental conditions.
+
+**Approach:** Define block correlation design, tune keepX per block via cross-validation, and fit the supervised DIABLO model.
+
+```r
+# DIABLO integrates multiple blocks with supervision
+# Finds features discriminating between conditions
+
+# Prepare block list
+X_blocks <- list(RNA = X_rna, Protein = X_protein)
+
+# Design matrix (correlation between blocks)
+design <- matrix(0.1, ncol = 2, nrow = 2, dimnames = list(names(X_blocks), names(X_blocks)))
+diag(design) <- 0
+
+# Tune parameters
+tune_diablo <- tune.block.splsda(X_blocks, Y, ncomp = 3,
+                                  test.keepX = list(RNA = c(10, 25, 50),
+                                                    Protein = c(10, 25, 50)),
+                                  design = design, validation = 'Mfold', folds = 5,
+                                  nrepeat = 10, cpus = 4)
+
+# Optimal keepX values
+optimal_keepX <- tune_diablo$choice.keepX
+
+# Final model
+diablo <- block.splsda(X_blocks, Y, ncomp = 3, keepX = optimal_keepX, design = design)
+
+# Performance
+perf_diablo <- perf(diablo, validation = 'Mfold', folds = 5, nrepeat = 10)
+plot(perf_diablo)
+```
+
+## DIABLO Visualization
+
+**Goal:** Visualize DIABLO results including sample separation, inter-block correlations, and feature networks.
+
+**Approach:** Use mixOmics plotting functions for consensus plots, circos plots, and correlation networks.
+
+```r
+# Sample plots
+plotIndiv(diablo, comp = c(1, 2), blocks = 'consensus', group = Y,
+          legend = TRUE, title = 'DIABLO Consensus')
+
+# Per-block sample plots
+plotIndiv(diablo, comp = c(1, 2), blocks = 'RNA', group = Y)
+
+# Variable plots
+plotVar(diablo, comp = c(1, 2), blocks = c('RNA', 'Protein'),
+        var.names = list(RNA = FALSE, Protein = FALSE))
+
+# Circos plot showing inter-block correlations
+circosPlot(diablo, cutoff = 0.7, line = TRUE)
+
+# Network of correlated features
+network(diablo, blocks = c('RNA', 'Protein'), cutoff = 0.6)
+
+# Heatmap
+cimDiablo(diablo, margin = c(8, 20))
+```
+
+## Extract Selected Features
+
+**Goal:** Retrieve the discriminant features selected by DIABLO for each omics block and export for pathway analysis.
+
+**Approach:** Use selectVar() to get selected variable names and plotLoadings() for contribution plots, then write to CSV.
+
+```r
+# Get selected variables per block
+selected_rna <- selectVar(diablo, block = 'RNA', comp = 1)$RNA$name
+selected_protein <- selectVar(diablo, block = 'Protein', comp = 1)$Protein$name
+
+# Loadings
+loadings_rna <- plotLoadings(diablo, block = 'RNA', comp = 1, contrib = 'max')
+loadings_protein <- plotLoadings(diablo, block = 'Protein', comp = 1, contrib = 'max')
+
+# Export for pathway analysis
+write.csv(data.frame(gene = selected_rna), 'diablo_rna_features.csv', row.names = FALSE)
+write.csv(data.frame(protein = selected_protein), 'diablo_protein_features.csv', row.names = FALSE)
+```
+
+## MINT: Multi-Study Integration
+
+**Goal:** Integrate data from multiple studies while accounting for study-specific batch effects.
+
+**Approach:** Fit MINT model with study indicator, then visualize global and per-study sample projections.
+
+```r
+# MINT for integrating multiple studies
+# Accounts for study-specific effects
+
+study <- factor(c(rep('Study1', 50), rep('Study2', 50)))
+
+mint_result <- mint.splsda(X = X_rna, Y = Y, study = study, ncomp = 3, keepX = c(50, 50, 50))
+
+# Visualize
+plotIndiv(mint_result, study = 'global', group = Y, legend = TRUE)
+plotIndiv(mint_result, study = 'all.partial', group = Y)
+
+# Performance
+perf_mint <- perf(mint_result, validation = 'Mfold', folds = 5)
+```
+
+## Unsupervised: sPCA and sPLS-DA Single Omics
+
+**Goal:** Perform sparse dimensionality reduction or classification on a single omics dataset.
+
+**Approach:** Apply sPCA for unsupervised exploration or sPLS-DA for supervised classification with feature selection.
+
+```r
+# Sparse PCA (single omics)
+spca_result <- spca(X_rna, ncomp = 3, keepX = c(50, 50, 50))
+plotIndiv(spca_result, group = Y)
+plotVar(spca_result)
+
+# sPLS-DA (single omics with supervision)
+splsda_result <- splsda(X_rna, Y, ncomp = 3, keepX = c(50, 50, 50))
+plotIndiv(splsda_result, group = Y, legend = TRUE)
+
+# Background prediction
+background <- background.predict(splsda_result, comp.predicted = 2, dist = 'max.dist')
+plotIndiv(splsda_result, group = Y, background = background)
+```
+
+## Model Performance and Validation
+
+**Goal:** Assess DIABLO classification performance via cross-validation error rates and AUC.
+
+**Approach:** Run repeated M-fold cross-validation and compute AUC per block and component.
+
+```r
+# Cross-validation performance
+perf_result <- perf(diablo, validation = 'Mfold', folds = 5, nrepeat = 50, cpus = 4)
+
+# Error rates
+plot(perf_result)
+perf_result$error.rate
+
+# AUC
+auc_diablo <- auroc(diablo, roc.block = 'RNA', roc.comp = 1)
+```
+
+## Related Skills
+
+- mofa-integration - Unsupervised multi-omics
+- data-harmonization - Preprocess before integration
+- differential-expression/de-results - Single-omics analysis
+- pathway-analysis/go-enrichment - Interpret selected features

package/data/skills/bio-multi-omics-mofa-integration/SKILL.md

@@ -0,0 +1,225 @@
+---
+name: bio-multi-omics-mofa-integration
+description: Multi-Omics Factor Analysis (MOFA2) for unsupervised integration of multiple data modalities. Identifies shared and view-specific sources of variation. Use when integrating RNA-seq, proteomics, methylation, or other omics to discover latent factors driving biological variation across modalities.
+tool_type: r
+primary_tool: MOFA2
+---
+
+## Version Compatibility
+
+Reference examples tested with: scanpy 1.10+
+
+Before using code patterns, verify installed versions match. If versions differ:
+- R: `packageVersion('<pkg>')` then `?function_name` to verify parameters
+
+If code throws ImportError, AttributeError, or TypeError, introspect the installed
+package and adapt the example to match the actual API rather than retrying.
+
+# MOFA2 Integration
+
+**"Find shared variation across my omics layers"** → Discover latent factors that capture shared and modality-specific sources of biological variation in an unsupervised manner.
+- R: `MOFA2::create_mofa()` → `prepare_mofa()` → `run_mofa()`
+- Python: `mofapy2` for training, `muon` for downstream
+
+## Prepare Multi-Omics Data
+
+**Goal:** Load and align multiple omics matrices into a consistent format for MOFA2 input.
+
+**Approach:** Read each omics layer, intersect to common samples, transpose to features-by-samples orientation.
+
+```r
+library(MOFA2)
+library(MultiAssayExperiment)
+
+# Load individual omics matrices (samples x features)
+rna <- as.matrix(read.csv('rnaseq_matrix.csv', row.names = 1))
+protein <- as.matrix(read.csv('proteomics_matrix.csv', row.names = 1))
+methylation <- as.matrix(read.csv('methylation_matrix.csv', row.names = 1))
+
+# Ensure consistent sample names across views
+common_samples <- Reduce(intersect, list(rownames(rna), rownames(protein), rownames(methylation)))
+rna <- rna[common_samples, ]
+protein <- protein[common_samples, ]
+methylation <- methylation[common_samples, ]
+
+# Transpose to features x samples (MOFA format)
+data_list <- list(
+    RNA = t(rna),
+    Protein = t(protein),
+    Methylation = t(methylation)
+)
+```
+
+## Create and Train MOFA Model
+
+**Goal:** Configure and train a MOFA2 model to discover shared and view-specific latent factors.
+
+**Approach:** Set model and training options, then run variational inference to learn factor decomposition.
+
+```r
+# Create MOFA object
+mofa <- create_mofa(data_list)
+
+# View data overview
+plot_data_overview(mofa)
+
+# Set model options
+model_opts <- get_default_model_options(mofa)
+model_opts$num_factors <- 15  # Number of factors to learn
+
+# Set training options
+train_opts <- get_default_training_options(mofa)
+train_opts$convergence_mode <- 'slow'
+train_opts$seed <- 42
+
+# Prepare and train
+mofa <- prepare_mofa(mofa, model_options = model_opts, training_options = train_opts)
+mofa <- run_mofa(mofa, outfile = 'mofa_model.hdf5')
+```
+
+## Analyze Factors
+
+**Goal:** Quantify how much variance each factor explains per omics view and extract factor scores and loadings.
+
+**Approach:** Plot variance decomposition and retrieve factor values (sample scores) and weights (feature loadings) as data frames.
+
+```r
+# Variance explained per factor per view
+plot_variance_explained(mofa, max_r2 = 15)
+plot_variance_explained(mofa, plot_total = TRUE)[[2]]
+
+# Factor values (sample scores)
+factors <- get_factors(mofa, as.data.frame = TRUE)
+
+# Factor weights (feature loadings)
+weights <- get_weights(mofa, as.data.frame = TRUE)
+```
+
+## Visualize Results
+
+**Goal:** Generate publication-quality plots of factor values, feature weights, and factor correlations.
+
+**Approach:** Use MOFA2 built-in plotting functions for scatter plots, heatmaps, and correlation matrices.
+
+```r
+# Scatter plot of factor values
+plot_factor(mofa, factor = 1, color_by = 'Group')
+
+# Heatmap of factor weights
+plot_weights(mofa, view = 'RNA', factor = 1, nfeatures = 20)
+
+# Top features per factor
+plot_top_weights(mofa, view = 'RNA', factor = 1, nfeatures = 10)
+
+# Correlation between factors
+plot_factor_cor(mofa)
+```
+
+## Factor Interpretation
+
+**Goal:** Identify biological pathways and gene sets associated with each MOFA factor.
+
+**Approach:** Extract top-weighted features per factor and run gene set enrichment on factor weights.
+
+```r
+# Extract top features for pathway analysis
+top_rna_factor1 <- get_weights(mofa, views = 'RNA', factors = 1, as.data.frame = TRUE)
+top_rna_factor1 <- top_rna_factor1[order(abs(top_rna_factor1$value), decreasing = TRUE), ]
+gene_list <- head(top_rna_factor1$feature, 100)
+
+# Gene set enrichment on factor weights
+library(MOFA2)
+enrichment <- run_enrichment(mofa, feature.sets = msigdb_genesets,
+                              view = 'RNA', factors = 1:5)
+plot_enrichment(enrichment, factor = 1, max.pathways = 15)
+```
+
+## Add Sample Metadata
+
+**Goal:** Annotate MOFA factors with clinical or experimental metadata for colored visualizations.
+
+**Approach:** Load sample annotations and attach to the MOFA object, then plot factors colored by metadata variables.
+
+```r
+# Load sample annotations
+metadata <- read.csv('sample_metadata.csv', row.names = 1)
+
+# Add to MOFA object
+samples_metadata(mofa) <- metadata[samples_names(mofa)[[1]], ]
+
+# Color by metadata
+plot_factor(mofa, factor = 1, color_by = 'Condition')
+plot_factors(mofa, factors = 1:3, color_by = 'Condition')
+```
+
+## Multi-Group MOFA
+
+**Goal:** Apply MOFA to batch- or group-structured multi-omics data and compare factor activity across groups.
+
+**Approach:** Organize data as nested list by group, train multi-group MOFA, and visualize group-specific factor patterns.
+
+```r
+# For batch/group-structured data
+data_list_grouped <- list(
+    group1 = list(RNA = rna_g1, Protein = prot_g1),
+    group2 = list(RNA = rna_g2, Protein = prot_g2)
+)
+
+mofa_grouped <- create_mofa(data_list_grouped)
+mofa_grouped <- prepare_mofa(mofa_grouped)
+mofa_grouped <- run_mofa(mofa_grouped)
+
+# Compare factor activity across groups
+plot_factor(mofa_grouped, factor = 1, group_by = 'group', color_by = 'group')
+```
+
+## MOFA+ for Single-Cell
+
+**Goal:** Apply MOFA to single-cell multi-modal data (CITE-seq, Multiome) with stochastic inference for scalability.
+
+**Approach:** Extract modality matrices from a Seurat object, create MOFA with stochastic training for large cell counts.
+
+```r
+# For single-cell multi-omics (CITE-seq, Multiome)
+library(Seurat)
+
+# Extract modalities from Seurat object
+rna_mat <- GetAssayData(seurat_obj, assay = 'RNA', layer = 'data')
+adt_mat <- GetAssayData(seurat_obj, assay = 'ADT', layer = 'data')
+
+# Create MOFA with single-cell settings
+mofa_sc <- create_mofa(list(RNA = rna_mat, ADT = adt_mat))
+model_opts <- get_default_model_options(mofa_sc)
+model_opts$num_factors <- 10
+
+# Use stochastic inference for large datasets
+train_opts <- get_default_training_options(mofa_sc)
+train_opts$stochastic <- TRUE
+```
+
+## Export Results
+
+**Goal:** Save MOFA factor scores, feature weights, and variance explained to CSV for downstream use.
+
+**Approach:** Extract each result type as a data frame and write to disk.
+
+```r
+# Save factor values
+factors_df <- get_factors(mofa, as.data.frame = TRUE)
+write.csv(factors_df, 'mofa_factors.csv', row.names = FALSE)
+
+# Save weights
+weights_df <- get_weights(mofa, as.data.frame = TRUE)
+write.csv(weights_df, 'mofa_weights.csv', row.names = FALSE)
+
+# Save variance explained
+var_exp <- get_variance_explained(mofa)
+write.csv(var_exp$r2_per_factor, 'mofa_variance_explained.csv')
+```
+
+## Related Skills
+
+- mixomics-analysis - Supervised multi-omics integration
+- data-harmonization - Preprocess data before MOFA
+- pathway-analysis/go-enrichment - Interpret MOFA factors
+- single-cell/multimodal-integration - Single-cell multi-omics

package/data/skills/bio-multi-omics-similarity-network/SKILL.md

@@ -0,0 +1,235 @@
+---
+name: bio-multi-omics-similarity-network
+description: Similarity Network Fusion (SNF) for patient stratification using multi-omics data. Integrates multiple data types into a unified patient similarity network. Use when performing patient stratification or integrating multi-omics data into unified similarity networks.
+tool_type: r
+primary_tool: SNFtool
+---
+
+## Version Compatibility
+
+Reference examples tested with: scanpy 1.10+
+
+Before using code patterns, verify installed versions match. If versions differ:
+- R: `packageVersion('<pkg>')` then `?function_name` to verify parameters
+
+If code throws ImportError, AttributeError, or TypeError, introspect the installed
+package and adapt the example to match the actual API rather than retrying.
+
+# Similarity Network Fusion
+
+**"Stratify patients using multi-omics data"** → Fuse omics-specific patient similarity networks into a unified network for subtype discovery and clustering.
+- R: `SNFtool::SNF()` to fuse networks, `spectralClustering()` for subtyping
+
+## Basic SNF Workflow
+
+**Goal:** Fuse multiple omics-specific patient similarity networks into a single unified network.
+
+**Approach:** Compute per-omics distance and affinity matrices, then iteratively fuse with SNF.
+
+```r
+library(SNFtool)
+
+# Load omics data (samples x features)
+data1 <- as.matrix(read.csv('rnaseq.csv', row.names = 1))
+data2 <- as.matrix(read.csv('methylation.csv', row.names = 1))
+data3 <- as.matrix(read.csv('mirna.csv', row.names = 1))
+
+# Ensure matching samples
+common <- Reduce(intersect, list(rownames(data1), rownames(data2), rownames(data3)))
+data1 <- data1[common, ]
+data2 <- data2[common, ]
+data3 <- data3[common, ]
+
+# Compute distance matrices
+dist1 <- dist2(as.matrix(data1), as.matrix(data1))
+dist2 <- dist2(as.matrix(data2), as.matrix(data2))
+dist3 <- dist2(as.matrix(data3), as.matrix(data3))
+
+# Construct affinity matrices
+# K = number of neighbors, alpha = hyperparameter
+K <- 20
+alpha <- 0.5
+
+aff1 <- affinityMatrix(dist1, K, alpha)
+aff2 <- affinityMatrix(dist2, K, alpha)
+aff3 <- affinityMatrix(dist3, K, alpha)
+
+# Fuse networks
+# T = number of iterations
+fused <- SNF(list(aff1, aff2, aff3), K = K, t = 20)
+```
+
+## Cluster Patients
+
+**Goal:** Identify patient subtypes from the fused similarity network using spectral clustering.
+
+**Approach:** Estimate optimal cluster count from the fused graph, then apply spectral clustering.
+
+```r
+# Determine optimal number of clusters
+estimateNumberOfClustersGivenGraph(fused, NUMC = 2:10)
+
+# Spectral clustering
+num_clusters <- 3
+clusters <- spectralClustering(fused, num_clusters)
+
+# Add to sample metadata
+sample_info <- data.frame(
+    Sample = rownames(data1),
+    Cluster = factor(clusters)
+)
+```
+
+## Visualize Network
+
+**Goal:** Display the fused patient network as a graph and heatmap with cluster annotations.
+
+**Approach:** Convert the fused matrix to an igraph object, filter weak edges, and render with cluster coloring.
+
+```r
+library(igraph)
+
+# Convert to igraph
+g <- graph_from_adjacency_matrix(fused, mode = 'undirected', weighted = TRUE, diag = FALSE)
+
+# Remove weak edges
+threshold <- quantile(E(g)$weight, 0.9)
+g_filtered <- delete_edges(g, E(g)[weight < threshold])
+
+# Plot
+V(g_filtered)$color <- clusters
+plot(g_filtered, vertex.size = 5, vertex.label = NA,
+     edge.width = E(g_filtered)$weight * 2,
+     main = 'SNF Patient Network')
+
+# Heatmap
+library(pheatmap)
+pheatmap(fused, cluster_rows = TRUE, cluster_cols = TRUE,
+         annotation_row = sample_info['Cluster'],
+         show_rownames = FALSE, show_colnames = FALSE)
+```
+
+## Normalized Mutual Information
+
+**Goal:** Evaluate clustering quality by comparing SNF clusters against known subtypes and single-omics baselines.
+
+**Approach:** Compute NMI between predicted clusters and true labels for fused vs individual affinity networks.
+
+```r
+# Compare with known labels
+true_labels <- read.csv('phenotype.csv')$Subtype
+
+# NMI score
+nmi <- calNMI(clusters, true_labels)
+cat('NMI:', nmi, '\n')
+
+# Compare individual vs fused
+nmi_rna <- calNMI(spectralClustering(aff1, num_clusters), true_labels)
+nmi_meth <- calNMI(spectralClustering(aff2, num_clusters), true_labels)
+nmi_mirna <- calNMI(spectralClustering(aff3, num_clusters), true_labels)
+
+cat('NMI RNA only:', nmi_rna, '\n')
+cat('NMI Methylation only:', nmi_meth, '\n')
+cat('NMI miRNA only:', nmi_mirna, '\n')
+cat('NMI Fused:', nmi, '\n')
+```
+
+## Feature Ranking with SNF
+
+**Goal:** Rank features by their contribution to the SNF-derived patient clusters.
+
+**Approach:** Perform ANOVA per feature across cluster assignments, ranking by F-statistic p-value.
+
+```r
+# Rank features by their contribution to clustering
+# Using network-based method
+
+# For each omics layer
+rank_features <- function(data, clusters) {
+    # Calculate feature importance based on cluster separation
+    f_values <- apply(data, 2, function(x) {
+        summary(aov(x ~ factor(clusters)))[[1]][1, 4]
+    })
+    f_values[is.na(f_values)] <- 1
+    names(sort(f_values))
+}
+
+top_rna <- rank_features(data1, clusters)
+top_meth <- rank_features(data2, clusters)
+```
+
+## Survival Analysis with Clusters
+
+**Goal:** Assess clinical relevance of SNF clusters by comparing survival outcomes between subtypes.
+
+**Approach:** Fit Kaplan-Meier curves per cluster and test significance with the log-rank test.
+
+```r
+library(survival)
+library(survminer)
+
+# Load survival data
+surv_data <- read.csv('survival.csv')
+surv_data$Cluster <- clusters[match(surv_data$Sample, rownames(data1))]
+
+# Kaplan-Meier
+fit <- survfit(Surv(Time, Event) ~ Cluster, data = surv_data)
+
+ggsurvplot(fit, data = surv_data, pval = TRUE,
+           risk.table = TRUE, palette = 'jco',
+           title = 'SNF Cluster Survival')
+
+# Log-rank test
+survdiff(Surv(Time, Event) ~ Cluster, data = surv_data)
+```
+
+## Parameter Tuning
+
+**Goal:** Optimize SNF hyperparameters (K neighbors, alpha) for best clustering performance.
+
+**Approach:** Grid search over K and alpha values, evaluating each combination by NMI against known labels.
+
+```r
+# Grid search over K and alpha
+K_range <- c(10, 20, 30)
+alpha_range <- c(0.3, 0.5, 0.8)
+
+results <- expand.grid(K = K_range, alpha = alpha_range, NMI = NA)
+
+for (i in 1:nrow(results)) {
+    aff1 <- affinityMatrix(dist1, results$K[i], results$alpha[i])
+    aff2 <- affinityMatrix(dist2, results$K[i], results$alpha[i])
+    aff3 <- affinityMatrix(dist3, results$K[i], results$alpha[i])
+
+    fused <- SNF(list(aff1, aff2, aff3), K = results$K[i], t = 20)
+    clusters <- spectralClustering(fused, num_clusters)
+    results$NMI[i] <- calNMI(clusters, true_labels)
+}
+
+best <- results[which.max(results$NMI), ]
+cat('Best parameters: K =', best$K, ', alpha =', best$alpha, '\n')
+```
+
+## Integration with Clinical Features
+
+**Goal:** Incorporate clinical variables as an additional data view in the SNF fusion.
+
+**Approach:** Encode clinical features numerically, compute a clinical affinity matrix, and include it in the SNF fusion step.
+
+```r
+# Add clinical features as another view
+clinical <- read.csv('clinical.csv', row.names = 1)
+clinical_numeric <- model.matrix(~ . - 1, data = clinical)
+
+dist_clinical <- dist2(clinical_numeric, clinical_numeric)
+aff_clinical <- affinityMatrix(dist_clinical, K, alpha)
+
+# Fuse all including clinical
+fused_with_clinical <- SNF(list(aff1, aff2, aff3, aff_clinical), K = K, t = 20)
+```
+
+## Related Skills
+
+- mofa-integration - Factor-based integration
+- mixomics-analysis - Supervised integration
+- single-cell/clustering - Single-cell clustering methods