Prediction Model Pooling, Selection and Performance Evaluation Across Multiply Imputed Datasets

Pooling, backward and forward selection of linear, logistic and Cox regression models in multiply imputed datasets. Backward and forward selection can be done from the pooled model using Rubin's Rules (RR), the D1, D2, D3, D4 and the median p-values method. This is also possible for Mixed models. The models can contain continuous, dichotomous, categorical and restricted cubic spline predictors and interaction terms between all these type of predictors. The stability of the models can be evaluated using (cluster) bootstrapping. The package further contains functions to pool model performance measures as ROC/AUC, Reclassification, R-squared, scaled Brier score, H&L test and calibration plots for logistic regression models. Internal validation can be done across multiply imputed datasets with cross-validation or bootstrapping. The adjusted intercept after shrinkage of pooled regression coefficients can be obtained. Backward and forward selection as part of internal validation is possible. A function to externally validate logistic prediction models in multiple imputed datasets is available and a function to compare models. For Cox models a strata variable can be included. Eekhout (2017) . Wiel (2009) . Marshall (2009) .

Data and Statistical Analyses after Multiple Imputation

Statistical Analyses and Pooling after Multiple Imputation. A large variety of repeated statistical analysis can be performed and finally pooled. Statistical analysis that are available are, among others, Levene's test, Odds and Risk Ratios, One sample proportions, difference between proportions and linear and logistic regression models. Functions can also be used in combination with the Pipe operator. More and more statistical analyses and pooling functions will be added over time. Heymans (2007) . Eekhout (2017) . Wiel (2009) . Marshall (2009) . Sidi (2021) . Lott (2018) . Grund (2021) .

Self-Controlled Case Series

Execute the self-controlled case series (SCCS) design using observational data in the OMOP Common Data Model. Extracts all necessary data from the database and transforms it to the format required for SCCS. Age and season can be modeled using splines assuming constant hazard within calendar months. Event-dependent censoring of the observation period can be corrected for. Many exposures can be included at once (MSCCS), with regularization on all coefficients except for the exposure of interest. Includes diagnostics for all major assumptions of the SCCS.

Parentage Assignment using Bi-Allelic Genetic Markers

Can be used for paternity and maternity assignment and outperforms conventional methods where closely related individuals occur in the pool of possible parents. The method compares the genotypes of offspring with any combination of potentials parents and scores the number of mismatches of these individuals at bi-allelic genetic markers (e.g. Single Nucleotide Polymorphisms). It elaborates on a prior exclusion method based on the Homozygous Opposite Test (HOT; Huisman 2017 ) by introducing the additional exclusion criterion HIPHOP (Homozygous Identical Parents, Heterozygous Offspring are Precluded; Cockburn et al., in revision). Potential parents are excluded if they have more mismatches than can be expected due to genotyping error and mutation, and thereby one can identify the true genetic parents and detect situations where one (or both) of the true parents is not sampled. Package 'hiphop' can deal with (a) the case where there is contextual information about parentage of the mother (i.e. a female has been seen to be involved in reproductive tasks such as nest building), but paternity is unknown (e.g. due to promiscuity), (b) where both parents need to be assigned, because there is no contextual information on which female laid eggs and which male fertilized them (e.g. polygynandrous mating system where multiple females and males deposit young in a common nest, or organisms with external fertilisation that breed in aggregations). For details: Cockburn, A., Penalba, J.V.,Jaccoud, D.,Kilian, A., Brouwer, L., Double, M.C., Margraf, N., Osmond, H.L., van de Pol, M. and Kruuk, L.E.B. (in revision). HIPHOP: improved paternity assignment among close relatives using a simple exclusion method for bi-allelic markers. Molecular Ecology Resources, DOI to be added upon acceptance.

Synthesizing Causal Evidence in a Distributed Research Network

Routines for combining causal effect estimates and study diagnostics across multiple data sites in a distributed study, without sharing patient-level data. Allows for normal and non-normal approximations of the data-site likelihood of the effect parameter.

Cyclic Coordinate Descent for Logistic, Poisson and Survival Analysis

This model fitting tool incorporates cyclic coordinate descent and majorization-minimization approaches to fit a variety of regression models found in large-scale observational healthcare data. Implementations focus on computational optimization and fine-scale parallelization to yield efficient inference in massive datasets. Please see: Suchard, Simpson, Zorych, Ryan and Madigan (2013) .

Climate Window Analysis

Contains functions to detect and visualise periods of climate sensitivity (climate windows) for a given biological response. Please see van de Pol et al. (2016) and Bailey and van de Pol (2016) for details.

JAR Dependencies for the 'DatabaseConnector' Package

Provides external JAR dependencies for the 'DatabaseConnector' package.

Generating Features for a Cohort

An R interface for generating features for a cohort using data in the Common Data Model. Features can be constructed using default or custom made feature definitions. Furthermore it's possible to aggregate features and get the summary statistics.

Tools for Type S (Sign) and Type M (Magnitude) Errors

Provides tools for working with Type S (Sign) and Type M (Magnitude) errors, as proposed in Gelman and Tuerlinckx (2000) and Gelman & Carlin (2014) . In addition to simply calculating the probability of Type S/M error, the package includes functions for calculating these errors across a variety of effect sizes for comparison, and recommended sample size given "tolerances" for Type S/M errors. To improve the speed of these calculations, closed forms solutions for the probability of a Type S/M error from Lu, Qiu, and Deng (2018) are implemented. As of 1.0.0, this includes support only for simple research designs. See the package vignette for a fuller exposition on how Type S/M errors arise in research, and how to analyze them using the type of design analysis proposed in the above papers.