Behind SEQopts() | The options | General Options | Weighting Options | Plot Options (km.curves = TRUE) | Special Cases | Internal Options

Run fsw() on ivreg() model object | Run fsw() on AER::ivreg() model object | Run fsw() on estimatr::iv_robust() model object | Run fsw() on fixest::feols() model object | Comparison with F-statistic from lfe package | Comparison with output from ivreg2

Run fsw() on ivreg() model object | Run fsw() on AER::ivreg() model object | Run fsw() on estimatr::iv_robust() model object | Run fsw() on fixest::feols() model object | Comparison with F-statistic from lfe package | Comparison with output from ivreg2

What has changed | Dataset IDs | Authentication | UKBiobank data has been curated | All data is now harmonised | LD reference panel is now harmonised | Instrument lists are up-to-date | dbSNP rs IDs | Everything is faster | What is new | Browse available datasets online | Chromosome and position | INDELs are retained | Multi-allelic variants are retained | More data | Error messages are more informative | Easier programmatic access to the database | Local LD operations | Access the data directly | Connect the data to different analytical tools | Key links | How to request new data

Introduction | MR methods | Sensitivity analyses | Heterogeneity statistics | Horizontal pleiotropy | Single SNP analysis | Leave-one-out analysis | Plots | Scatter plot | Forest plot | Forest plot with categories | RadialMR outlier example | MR-PRESSO outlier example | Leave-one-out plot | Funnel plot | 1-to-many forest plot | Step 1: Generate 1-to-many MR results | Step 2: Make the 1-to-many forest plot | Example 1. Effect of multiple risk factors on coronary heart disease | Example 2. MR results for multiple MR methods grouped by multiple exposures | Example 3. Stratify results on a grouping variable | Example 4. Effect of BMI on 103 diseases | MR-RAPS: Many weak instruments analysis | MR-GRIP | Reports | MR Steiger directionality test | Multivariable MR | Note about multivariable methods | Note about visualisation | Using your own summary data | From local files | From data frames | Mixing local and OpenGWAS data | Converting to MVMR format | MR estimates when instruments are correlated | MR-MoE: Using a mixture of experts machine learning approach | Post MR results management | Split outcome names | Split exposure names | Generate odds ratios with 95% confidence intervals | Subset on method | Combine all results | References

What this is for | A minimal example | Missing measurements and the time window | Checking what contributed | Continuous outcomes | Per-protocol effects

Per-protocol, censoring, weights in pre-expanded data and no truncation, no excused conditions (i.e. static interventions) | Per-protocol, censoring, weights in post-expanded data and no truncation, no excused conditions (i.e. static interventions) | Per-protocol, censoring, weights in pre-expanded data and no truncation, excused conditions for initiators and non-initiators (i.e. dynamic interventions) | Per-protocol, censoring, a flexible baseline hazard in the weight models | Per-protocol, censoring, weights in post-expanded data and no truncation, excused conditions for initiators and non-initiators (i.e. dynamic interventions) | Per-protocol, censoring, weights in post-expanded data and no truncation, excused conditions for initiators and non-initiators (i.e. dynamic interventions) and a competing event | Per-protocol, censoring, weights in post-expanded data and no truncation, excused conditions for initiators and non-initiators (i.e. dynamic interventions) and hazard ratio

1. Regular VS Code settings | 2. Additional VS Code settings | 3. Launch Config | 3.1 Launch Requests | 3.2 Attach Requests | 3.3 Shared config entries | 4. R Options

First attach the Statamarkdown library | If Stata was not found | Then set up Stata code chunks.

Some Setup for Stata | Using Stata | Using R

Setup Stata Engine | An Example | First code block: | Second code block:

First attach the Statamarkdown library | If Stata was not found | Then set up Stata code chunks.

Some Setup for Stata | Using Stata | Using R

Setup Stata Engine | An Example | First code block: | Second code block:

ITT With 5 bootstrap samples | ITT with 5 bootstrap samples and losses-to-followup | ITT with 5 bootstrap samples and competing events | ITT hazard ratio with 5 bootstrap samples and competing events | ITT with 5 bootstrap samples and competing events in subgroups defined by sex

Setting up your Analysis | Step 1 - Defining your options | Step 2 - Running the Primary Function | Step 3 - Recovering your results

Dose-response With 5 bootstrap samples | Dose-response with 5 bootstrap samples and losses-to-followup | Dose-response with 5 bootstrap samples and competing events | Dose-response hazard ratio with 5 bootstrap samples and competing events | Dose-response with 5 bootstrap samples and competing events in subgroups defined by sex

Comparison fits

Comparison fits

Introduction | Reading in from a file | Example 1: The default column names are used | Example 2: The text file has non-default column names | Using an existing data frame | Obtaining instruments from existing catalogues | GWAS catalog | Metabolites | Proteins | Gene expression levels | DNA methylation levels | IEU OpenGWAS database | Clumping

Introduction | Dealing with strand issues | Correct, unambiguous | Incorrect reference, unambiguous | Ambiguous | Palindromic SNP, inferrable | Palindromic SNP, not inferrable | Options | Drop duplicate exposure-outcome summary sets

Background | Installation | Overview | Authentication | References

Available studies in IEU GWAS database | Extracting particular SNPs from particular studies | LD proxies | Using local GWAS summary data | Outcome data format | More advanced use of local data

Introduction | A simple example Rmd file | Parameterising our Rmd file | Parameterising a Quarto document | Customising the RStudio Knit button

Introduction | Features of the bpbounds package | Vitamin A supplementation example | Entering the data as conditional probabilities | Treating the data as bivariate | Mendelian randomization example | Simulated example that does not satisfy the IV conditions | Conclusion | References

Introduction | Features of the bpbounds package | Vitamin A supplementation example | Entering the data as conditional probabilities | Treating the data as bivariate | Mendelian randomization example | Simulated example that does not satisfy the IV conditions | Conclusion | References

Covariance and Phenotypic correlations in MVMR | References

Overview | Workflow | Step 1: Obtain summary data | Estimating pairwise covariances between SNP associations | Step 2: Format summary data | Step 3: Test for weak instruments | Step 4: Test for horizontal pleiotropy using conventional Q-statistic estimation | Step 5: Estimate causal effects | Step 6: Robust causal effect estimation. | References

Authentication | Deprecated Google authentication | Allowance | General API queries | Get API status | Get list of all available studies | Get list of a specific study | Extract particular associations from particular studies | Get the tophits from a study | Perform PheWAS | LD clumping | LD matrix | Variant information | Extracting GWAS summary data based on gene region | 1000 genomes annotations

Authentication | Deprecated Google authentication | Allowance | General API queries | Get API status | Get list of all available studies | Get list of a specific study | Extract particular associations from particular studies | Get the tophits from a study | Perform PheWAS | LD clumping | LD matrix | Variant information | Extracting GWAS summary data based on gene region | 1000 genomes annotations

Example of computation time | Smaller datasets | n = 100,000 | n = 10,000 | Varying the number of coefficients | n = 100,000, p = 5 | n = 1,000,000, p = 20 | Session info

Setup | Fitting the model | Standard errors | Heteroskedasticity-consistent (HC) standard errors | Cluster-robust standard errors | Covariance matrices directly | Note | Regression tables with gtsummary | Logistic regression | Robust standard errors in a gtsummary table

User Interface | Observational Data | Weight Models | Censoring due to treatment switching | Other informative censoring | Calculate Weights | Specify Outcome Model | Expand Trials | Set Expansion Options | Create Sequence of Trials Data | Sample or Load from Expanded Data | Fit Marginal Structural Model | Inference | Flowchart

Keyboard shortcuts

Overview | Get state | From R | From HTTP | From WebSockets | Get Renderers | Render plot | Remove plots | Get static IDs | Security

System requirements

Basic usage | Build the image | Run the container | Start the device | Advanced usage | Set defaults in Rprofile

Usage | Troubleshooting

Required Data | All-in-one analysis | Flexible Analysis

LD clumping | LD matrix | LD proxies

LD clumping | LD matrix | LD proxies

Introduction | Model fitters | Classes and Slots | User Constructor | Methods | fit_weights_model | fit_outcome_model | predict | Data Stores | show | save_expanded_data | read_expanded_data | sample_expanded_data

Introduction | Quick demo | Key features | Regression categories | Model selection strategies | Selection metrics | Multicollinearity | StepReg output | Use cases | Linear regression with the mtcars dataset | Example1: single dependent variable ("mpg") | Example2: multivariate regression ("mpg" and "drat") | Logistic regression with the remission dataset | Example1: using "forward" strategy | Example2: using "subset" strategy | Cox regression with the lung dataset | Poisson regression with the creditCard dataset | Interactive app | Session info

Features | Installation | Using the Debugger | Launch Mode | Attach Mode | Configuration | How it works | Debugging R Packages | Troubleshooting | Contributing

Setting breakpoints during debugging | Debug modes | Breakpoints in packages