Analysis of T cell receptor (TCR) repertoires | Install and load the current version of ClustIRR | Familiarize with TCR data | Familiarize with gene expression data | Using ClustIRR to compare the two repertoires, CMV vs. Flu | Step 1 Building a joint graph | Using igraph functions to inspect $J$ | Step 2 Detecting communities in $J$ | Step 3 Finding differentially abundant communities (DACs) | Posterior predictive checks | Taking a peek at the relative community abundances ($\beta$) | Inspecting DACs of interest | Show the CDR3$\alpha$\beta$ sequence logos of community 9 | Visualize the graph of community 9 | Integration of TCR and GEX data

Introduction | Installation | System requirements | ClustIRR algorithm | Input | Algorithm | Step 1. Compute TCR clonotype similarities in a repertoire with cluster_irr | Step 2-3. Construct TCR repertoire graphs and join them into $J$ | Run steps 1-3 with clustirr | Inspect the content of clust_irrs | Inspect graphs with plot_graph | You can evaluate $J$ with igraph | Step 4. community detection with detect_communities | Inspecting the outputs of detect_communities | Qualitative similarity between community abundance vectors with get_honeycombs | Quantitative similarity between community abundance vectors with get_cosine_similarity | Summary of communities is provided in the detect_communities outputs | Special functions: decoding communities with decode_community | Step 5. differential community occupancy (DCO) with dco | Step 6. Inspect results | Visualizing the distribution of $\beta$ with get_beta_violin_ag | Posterior predictive checks | Differential community abundance results $\rightarrow$ par. $\delta$ | Conclusion: you can also use custom community occupancy matrix for DCO!

Create experimental ``artifact'' -> merge C2 and C6 to create a bimodal velocity distribution | Session Info

Introduction | Installation | Data Structure | Required Columns | Important: The offset Column | Example Data | Exploratory Data Analysis | Raw Cell Velocities | Mean Velocity per Well | Model Fitting | Control Parameters | Interpreting Results | Overall Treatment Effects ($\delta_t$) | Visualizing Effects | From Log-Fold-Change to Fold-Change | Pairwise Comparisons | Comparison Matrix | Visualize $\rho$ | Visualize $\pi$ | Visualize $\rho$ vs. $\pi$ with a volcano plot | Violin Plots for Specific Comparisons | Model Diagnostics and Calibration | Posterior Predictive Checks (PPC) | Cell-Level Check | Well-Level Check | Leave-One-Out (LOO) diagnostics | MCMC diagnostics | Check $\hat{R}$ and ESS | Check for divergent transitions and other issues | Variance Components | Advanced: Dose-Response Profiles | Session Info

Introduction | Partially Generative Simulation | Experimental Design | Model Parameters | Plate Effects ($\alpha_p$) | Control Treatment (Offset) | Cell Velocity Distribution ($\kappa_w$) | Generating the Data | Inspecting the Simulated Data | Visualizing Simulated Data | Analyzing Simulated Data | Validation: Truth vs. Inference | Treatment Effects ($\delta_t$) | Variability Parameters | Fully Generative Simulation | Model Specification | Generate Data | Comparing Simulation Modes | Power Analysis | Simulation Loop | Interpreting Results | Session Info

Background | Data^1 | Data processing | scBubbletree workflow | Determine the clustering resolution (step 1) | Determining the resolution parameter $r$ for Louvain clustering | Determining the number of clusters $k$ for k-means clustering [alternative] | Clustering (step 2) and hierarchical grouping (step 3) of bubbles | Clustering with Louvain | Clustering with k-means ([alternative]) | Comparison of bubbletree based on Louvain and k-means clustering | Visual comparison of two bubbletrees | Visualization (step 4) | Attaching categorical features | Gini impurity index | Attaching numeric features | Quality control with r Biocpkg("scBubbletree") | scBubbletree can incorporate results from other clustering approaches | Adding custom visuals: cell-cell communication | Visualizing features of individual cells | Summary | Session Info

Introduction | Input | Model | Case Study A: analyzing IRRs | DGU analysis | Output format | Model checking | MCMC sampling | PPC: posterior predictive checks | PPCs: repertoire-specific | PPCs: overall | Results | DGU: differential gene usage | Promising hits | Promising hits [count] | GU: gene usage summary | Leave-one-out (LOO) analysis | LOO-DGU: variability of effect size $\gamma$ | LOO-DGU: variability of $\pi$ | LOO-GU: variability of the gene usage | Case Study B: analyzing IRRs containing biological replicates | Modeling | Posterior predictive checks | Analysis of estimated effect sizes | Session