Introduction | Installation | The round-trip | The storage layout | Advanced: writing and attaching resources directly | Streaming a flat resource from a cursor | Operations run as SQL | Genomic coordinates are preserved | Single-cell data | The filter, realize, analyze pattern | Multi-omics | When to use BiocDuckDB | Session information

Introduction | Installation | Quick start | Working with a DuckDBMatrix | Writing a matrix to the DuckDB backend | Sizing partitions for large or sparse writes | Subsetting | Matrix statistics | Sparse data | Lazy evaluation | Realizing large results in blocks | N-dimensional arrays | Introspecting a Parquet file you did not write | When to use DuckDBArray | Session information

Introduction | Installation | Lazy spatial columns | Spatial predicates and filtering | Layer-level engines for coordinate columns | GeoParquet I/O | When to use DuckDBSpatial | Session information

Scope | Architecture | From sf generics to spatial SQL | Two dispatch paths | GeoParquet I/O | Extending, and the BiocDuckDB integration | Session information

Introduction | What BiocDuckDB optimizes | A small, live comparison | Benchmark setup | Results | Summary | When this matters | Running your own benchmarks | Session information

Introduction | What DuckDBGRanges optimizes | A small, live comparison | Benchmark setup | Results | Summary | Backend comparison | When to use each | Running your own benchmarks | Session information

Installation | Citing MultiAssaySpatialExperiment | Why another spatial data structure? | When to use MASE | Prerequisites | A worked example | Building the object | Looking at what it holds | Subsetting keeps the links consistent | Annotating and aggregating | Handing the data to other tools | See also | Session info

Introduction | Building MASE objects | Optional: lazy Parquet I/O via BiocDuckDB | Minimal MASE | Adding spatial coordinates | Adding spatial shapes | Building from multiple SingleCellExperiment objects | prepMASE and buildSpatialMap | Validation checklist | Subsetting operations | Overview of subsetting operations | Basic subsetting | Subset by specimen | Subset by assay column | Bracket notation | Spatial subsetting | Subset by bounding box | Subset by polygon | Select data from multiple non-contiguous regions | Buffer-based subsets | Spatial annotation and aggregation | Example setup | Annotate with regions | Aggregate by region | Count points per region | Sum expression per region | Mean expression per region | Lower-level spatial joins | Aggregation strategies | Visualizing aggregated data | Labels ↔ shapes interoperability | Rasterization: shapes → labels | Vectorization: labels → shapes | Use cases | See also | Session info

Introduction | Part 1: Data import | Reading Xenium data | Basic usage | Options | When to load transcripts | Reading Visium data | Visium HD support | Reading CosMx data | FOV (Field of View) handling | Reading MERSCOPE data (MERFISH) | Choosing a reader | Converting from other Bioconductor classes | From SpatialExperiment | From SpatialFeatureExperiment | Reading multiple samples | Part 2: Real-world workflows | Workflow 1: Xenium + Visium integration | Background | Data preparation | Build MASE object | Exploratory analysis | Do the two assays overlap in space? | Which genes are shared between assays | Spatial annotation: Cell types in Visium spots | Reading the composition result | Spatial aggregation: Summarize Xenium by regions | Comparing expression between the two platforms | Integration insights | Workflow 2: Coordinate transformation and alignment | Converting Visium grid to microns | Applying affine transformations | Landmark-based alignment | Storing transformed coordinates | Checking the alignment visually | Workflow 3: Multi-sample analysis | Workflow 4: Iterative analysis pattern | Summary | See also | Session info

Scope | Anatomy of a MultiAssaySpatialExperiment | What it inherits from MultiAssayExperiment | The spatial layers it adds | The tables that tie observations to space | The relational schema | Session info

Summary of the MultiAssaySpatialExperiment API | Constructors | Platform readers | Spatial accessors | Inherited accessors used most often | Subsetting | Spatial operations | Coercion | Combining | Session info

Introduction | Installation | Quick start | Working with a DuckDBGRanges | Accessors | Subsetting | Range operations | Filter, then materialize | Grouped ranges: DuckDBGRangesList | When to use DuckDBGRanges | Session information

Introduction | What DuckDBArray optimizes | A small, live comparison | Benchmark setup | Giving every backend its best effort | Results | Main takeaways | Backend comparison | When to use each | Running your own benchmarks | Session information

Scope | Architecture | Slots | The five required columns | From range operations to SQL | Grouped ranges as LIST columns | Materialization and interoperability | Session information

Scope | Architecture | The seed contract | Shape: dim() and dimnames() | Dense blocks: extract_array() | Sparse blocks: extract_sparse_array() | Wrapping the seed | From array operations to SQL | Grid partitioning | Realization and interoperability | Session information

Introduction | Relationship to arrow | Installation | Quick start | Working with a DuckDBDataFrame | Column access | Row subsetting | Computed columns | Column metadata | Columns come in three flavors | Reaching for SQL directly | When to use DuckDBDataFrame | Session information

Scope | The classes | The DuckDBTable abstraction | Construction | The contract | Key-dimension semantics | From R to SQL | Connection management | Dimension tables | Extending DuckDBDataFrame | Session information

Simulating trial data | Simulating external control data | Merging trial and external control data | Propensity scores overlap | Matching Methods | PSML | PSMR | GM | GMW | OM | Weighting Methods | GBM | EB | IPW | Balance Plots for Matching Methods | Balance Plots for Weighting Methods | Selection of matching/weighting methods | References

Before you start | Creating an analysis object | 1. data_matrix | Required elements | Time-to-event | Binary endpoints | Covariates | Example data | Creating a data matrix with create_data_matrix() | psborrow2 example matrix | 2. outcome | 3. borrowing | 3. treatment | Application | Sampling from an analysis object | Summarizing results | Using bayesplot | Using posterior

Hazard function | Cox proportional hazards model | Baseline hazard function | Likelihood | Power prior model | Commensurate prior model | Conducting Analysis | Example data | Exponential distribution with constant hazard with gamma prior distribution (Using psborrow2 R package) | Time-to-event analysis with Weibull distribution and proportional hazards parametrization | Power prior with piecewise exponential distribution (not implemented in psborrow2 R package) | Commensurate prior with piecewise exponential distribution (not implemented in psborrow2 R package) | Exponential distribution (constant hazard) and no borrowing: Using psborrow2 R package | Exponential distribution (constant hazard) and full borrowing: Using psborrow2 R package | Cox proportional hazards model using frequentist approach (No borrowing) | Cox proportional hazards model using frequentist approach (Full borrowing from external control) | Summarizing all the results together | References

Bringing your own simulated data | data_matrix_list | data_list | guide | effect, drift, and index | outcome | borrowing | covariate | treatment | create_simulation_obj() | mcmc_sample() | MSE | Type I error | Power | EHSS | Optimal Accuracy Design | References

Alternative PS Weights with WeightIt | Matching with MatchIt | Combined Weighting and Dynamic Borrowing | Fixed Weights | Reference Models | Comparison of Results | References

Introduction | Logistic regression | glm | BayesPPD | psborrow2 | Results | Exponential models

Generating Baseline Data | Generating Survival Data | Enrollment | Drop Out | Clinical Cut Off | Running a simulation | Using fixed external data | Combining simulations

Purpose | Explore example data | {psborrow2} contains an example matrix | Load as data.frame for some functions | Look at distribution of arms | Naive internal comparisons | Kaplan-Meier curves | Cox model | Hybrid control analysis | A note on prior distributions | Outcome objects | Create an exponential survival distribution Outcome object | Borrowing object | Treatment objects | Analysis objects | MCMC sampling | Interpret results | Control arm imbalances | Propensity score analysis with BDB

Specifying prior distributions | Types of prior distributions | Visualizing prior distributions

0. Analysis workflow & supported model types | 1. ER model development | 2. Simulation from developed ER model | 3. Plot simulated ER curve | Acknowledgement

Introduction | 1. Apply Bayesian dynamic borrowing methods | 2. Conduct simulation studies of Bayesian dynamic borrowing methods | 3. Generate data for simulation studies | Additional articles | Installing cmdstanr | References