Skip to contents

Screening predictors with MakeUnivariateRegressionTable()

Source: vignettes/UnivariateRegressions.qmd

1 Overview

MakeUnivariateRegressionTable() performs large numbers of regression analyses simultaneously and returns publication-ready summary tables.

This approach is particularly useful for:

  • Biomarker screening
  • Variable prioritization
  • Covariate discovery
  • Exploratory analyses
  • Hypothesis generation

This vignette demonstrates:

  • Running multiple univariate regressions
  • Publication-ready tables
  • Detailed regression tables
  • Covariate adjustment
  • Standardized coefficients
  • Forest plot visualization

2 Load packages

3 Load example data 

data("SampleData")
data("SampleVariableTypes")

RevaluedObj <- RevalueData(
  SampleData,
  SampleVariableTypes
)

df_Revalued <- RevaluedObj$RevaluedData

4 Define outcomes and predictors

For this example we will evaluate clinical and demographic predictors of continuous biomarker outcomes.

OutcomeVars <- c(
  "Calbindin",
  "Ferritin"
)

PredictorVars <- c(
  "Diagnosis",
  "age",
  "sex",
  "Genotype"
)

5 Create regression tables 

UniObj <- MakeUnivariateRegressionTable(
  data = df_Revalued,
  outcome_vars = OutcomeVars,
  predictor_vars = PredictorVars
)

6 Publication-ready table

The formatted table is designed for reporting and manuscript preparation.

UniObj$FormattedTable
Calbindin
Ferritin
Estimate (95% CI) p-value Estimate (95% CI) p-value
Diagnosis : Impaired 0.936 (-0.075, 1.95) 0.069 0.203 (0.0147, 0.392)* 0.035
Age 0.0164 (-0.0185, 0.0513) 0.36 0.0018 (-0.00474, 0.00834) 0.59
Sex : Male -1.62 (-2.53, -0.704)*** <0.001 0.253 (0.0815, 0.425)** 0.004
Genotype : E2E3 1.74 (-4.26, 7.75) 0.57 0.218 (-0.907, 1.34) 0.7
Genotype : E2E4 3.13 (-3.42, 9.67) 0.35 0.223 ( -1, 1.45) 0.72
Genotype : E3E3 1.81 (-4.08, 7.69) 0.55 0.195 (-0.907, 1.3) 0.73
Genotype : E3E4 2.46 (-3.44, 8.37) 0.41 0.267 (-0.838, 1.37) 0.63
Genotype : E4E4 1.31 (-4.98, 7.6) 0.68 -0.059 (-1.24, 1.12) 0.92

Each outcome has its own column spanner. Effects are displayed as estimate (95% CI), with significance stars appended and both the effect and p-value bolded when significant.

7 Detailed regression table

The larger table keeps the same wide outcome layout while exposing N, estimate, standard error, confidence limits, and p-value separately.

UniObj$LargeTable
Calbindin
Ferritin
N Estimate SE 95% CI Low 95% CI High p-value N Estimate SE 95% CI Low 95% CI High p-value
Diagnosis : Impaired 333.000 0.936 0.514 −0.075 1.948 0.069 333.000 0.203 0.096 0.015 0.392 0.035
Age 322.000 0.016 0.018 −0.018 0.051 0.355 322.000 0.002 0.003 −0.005 0.008 0.589
Sex : Male 333.000 −1.617 0.464 −2.530 −0.704 0.001 333.000 0.253 0.087 0.082 0.425 0.004
Genotype : E2E3 333.000 1.745 3.054 −4.263 7.753 0.568 333.000 0.218 0.572 −0.907 1.343 0.703
Genotype : E2E4 333.000 3.127 3.326 −3.415 9.670 0.348 333.000 0.223 0.623 −1.001 1.448 0.720
Genotype : E3E3 333.000 1.807 2.992 −4.080 7.694 0.546 333.000 0.195 0.560 −0.907 1.297 0.728
Genotype : E3E4 333.000 2.464 3.003 −3.443 8.371 0.412 333.000 0.267 0.562 −0.838 1.373 0.635
Genotype : E4E4 333.000 1.310 3.195 −4.976 7.596 0.682 333.000 −0.059 0.598 −1.236 1.118 0.921

This version is often useful during exploratory analyses and quality control.

8 Including covariates

Covariates can be included in every regression model.

For example, MMP7 can be included when evaluating whether associations are independent of another biomarker.

UniObj_Covar <- MakeUnivariateRegressionTable(
  data = df_Revalued,
  outcome_vars = OutcomeVars,
  predictor_vars = PredictorVars,
  covariates = "MMP7"
)
UniObj_Covar$FormattedTable
Calbindin
Ferritin
Estimate (95% CI) p-value Estimate (95% CI) p-value
Diagnosis : Impaired 1.05 (0.00979, 2.1)* 0.048 0.144 (-0.0493, 0.337) 0.14
Age 0.0161 (-0.0189, 0.051) 0.37 0.00222 (-0.00425, 0.0087) 0.5
Sex : Male -1.62 (-2.54, -0.696)*** <0.001 0.223 (0.0513, 0.395)* 0.011
Genotype : E2E3 1.73 (-4.29, 7.74) 0.57 0.243 (-0.87, 1.36) 0.67
Genotype : E2E4 3.09 (-3.46, 9.65) 0.35 0.272 (-0.94, 1.48) 0.66
Genotype : E3E3 1.79 (-4.11, 7.68) 0.55 0.221 (-0.869, 1.31) 0.69
Genotype : E3E4 2.44 (-3.47, 8.36) 0.42 0.294 (-0.8, 1.39) 0.6
Genotype : E4E4 1.28 (-5.01, 7.58) 0.69 -0.02 (-1.18, 1.14) 0.97

This allows investigators to determine whether associations remain significant after accounting for potential confounding factors.

9 Standardized coefficients

When predictors are measured on different scales, standardized coefficients can simplify interpretation.

UniObj_Std <- MakeUnivariateRegressionTable(
  data = df_Revalued,
  outcome_vars = OutcomeVars,
  predictor_vars = PredictorVars,
  Standardize = TRUE
)
UniObj_Std$FormattedTable
Calbindin
Ferritin
Estimate (95% CI) p-value Estimate (95% CI) p-value
Diagnosis : Impaired 0.223 (-0.0179, 0.464) 0.069 0.259 (0.0187, 0.5)* 0.035
Age 0.0517 (-0.0581, 0.161) 0.36 0.0303 (-0.0799, 0.14) 0.59
Sex : Male -0.385 (-0.603, -0.168)*** <0.001 0.323 (0.104, 0.542)** 0.004
Genotype : E2E3 0.416 (-1.02, 1.85) 0.57 0.278 (-1.16, 1.71) 0.7
Genotype : E2E4 0.745 (-0.814, 2.31) 0.35 0.285 (-1.28, 1.85) 0.72
Genotype : E3E3 0.431 (-0.973, 1.83) 0.55 0.248 (-1.16, 1.65) 0.73
Genotype : E3E4 0.587 (-0.821, 2) 0.41 0.341 (-1.07, 1.75) 0.63
Genotype : E4E4 0.312 (-1.19, 1.81) 0.68 -0.0753 (-1.58, 1.43) 0.92

Standardized coefficients represent changes in standard deviation units and facilitate comparison across predictors.

10 Accessing model objects

The fitted regression models are returned in the output object.

names(
  UniObj$ModelSummaries
)
NULL

For example:

summary(
  UniObj$ModelSummaries$Calbindin$Diagnosis
)
Length  Class   Mode
     0   NULL   NULL

This allows additional diagnostics and custom analyses.

11 Creating a forest plot

A forest plot provides a compact visual summary of regression results.

ForestPlot <- PlotForestFromTable(
  UniObj
)

ForestPlot

Forest plots make it easy to identify:

  • Strong predictors
  • Significant predictors
  • Direction of effects
  • Precision of estimates

12 Forest plot with covariate adjustment

The same visualization can be generated using adjusted models.

PlotForestFromTable(
  UniObj_Covar
)

Comparing adjusted and unadjusted models can help identify associations that may be explained by confounding variables.

A common workflow is:

  1. Use MakeUnivariateRegressionTable() to screen large numbers of predictors.
  2. Review the formatted table.
  3. Compare standardized and unstandardized results.
  4. Add important covariates.
  5. Visualize findings using PlotForestFromTable().
  6. Follow up significant findings using multivariable models.

14 Summary

MakeUnivariateRegressionTable() provides a rapid screening framework for evaluating many predictors across multiple outcomes.

Key features include:

  • Multiple outcomes
  • Multiple predictors
  • Optional covariate adjustment
  • Standardized coefficients
  • Publication-ready tables
  • Detailed regression output
  • Forest plot visualization

16 Session information 

R version 4.6.1 (2026-06-24)
Platform: x86_64-pc-linux-gnu
Running under: Ubuntu 24.04.4 LTS

Matrix products: default
BLAS:   /usr/lib/x86_64-linux-gnu/openblas-pthread/libblas.so.3
LAPACK: /usr/lib/x86_64-linux-gnu/openblas-pthread/libopenblasp-r0.3.26.so;  LAPACK version 3.12.0

locale:
 [1] LC_CTYPE=C.UTF-8       LC_NUMERIC=C           LC_TIME=C.UTF-8
 [4] LC_COLLATE=C.UTF-8     LC_MONETARY=C.UTF-8    LC_MESSAGES=C.UTF-8
 [7] LC_PAPER=C.UTF-8       LC_NAME=C              LC_ADDRESS=C
[10] LC_TELEPHONE=C         LC_MEASUREMENT=C.UTF-8 LC_IDENTIFICATION=C

time zone: UTC
tzcode source: system (glibc)

attached base packages:
[1] stats     graphics  grDevices utils     datasets  methods   base

other attached packages:
[1] dplyr_1.2.1           SciDataReportR_21.0.0

loaded via a namespace (and not attached):
 [1] gtable_0.3.6           xfun_0.60              bayestestR_0.18.1
 [4] ggplot2_4.0.3          insight_1.5.2          rstatix_1.1.0
 [7] lattice_0.22-9         paletteer_1.7.0        vctrs_0.7.3
[10] tools_4.6.1            generics_0.1.4         datawizard_1.3.1
[13] tibble_3.3.1           pkgconfig_2.0.3        RColorBrewer_1.1-3
[16] correlation_0.8.8      S7_0.2.2               RcppParallel_6.2.0
[19] gt_1.3.0               lifecycle_1.0.5        compiler_4.6.1
[22] farver_2.1.2           carData_3.0-6          snakecase_0.11.1
[25] sass_0.4.10            htmltools_0.5.9        yaml_2.3.12
[28] Formula_1.2-6          pillar_1.11.1          car_3.1-5
[31] tidyr_1.3.2            statsExpressions_2.0.0 abind_1.4-8
[34] tidyselect_1.2.1       sjlabelled_1.2.0       digest_0.6.39
[37] mvtnorm_1.4-2          gtsummary_2.5.1        purrr_1.2.2
[40] rematch2_2.1.2         labeling_0.4.3         forcats_1.0.1
[43] ggstatsplot_1.0.0      labelled_2.16.0        fastmap_1.2.0
[46] grid_4.6.1             cli_3.6.6              magrittr_2.0.5
[49] patchwork_1.3.2        dichromat_2.0-1        broom_1.0.13
[52] withr_3.0.3            scales_1.4.0           backports_1.5.1
[55] estimability_2.0.0     rmarkdown_2.31         emmeans_2.0.4
[58] otel_0.2.0             hms_1.1.4              coda_0.19-4.1
[61] evaluate_1.0.5         knitr_1.51             haven_2.5.5
[64] parameters_0.29.2      rstantools_2.7.0       rlang_1.3.0
[67] xtable_1.8-8           glue_1.8.1             xml2_1.6.0
[70] jsonlite_2.0.0         effectsize_1.0.3       R6_2.6.1
[73] fs_2.1.0