The Multivariate Asymptotic Non-parametric Test of Association (MANTA) enables non-parametric, asymptotic P-value computation for multivariate linear models.
# install.packages("devtools")
::install_github("dgarrimar/manta") devtools
R 3.3.2 or higher is required.
library(manta)
manta(biomarkers ~ ., data = patients)
#>
#> Call:
#> manta(formula = biomarkers ~ ., data = patients)
#>
#> Type II Sum of Squares
#>
#> Df Sum Sq Mean Sq F value R2 Pr(>F)
#> age 1 400.6 400.63 7.3566 0.04242 0.001685 **
#> gender 1 34.3 34.28 0.6295 0.00363 0.5144
#> status 2 2152.7 1076.33 19.7643 0.22793 1.348e-12 ***
#> Residuals 91 4955.7 54.46
#> ---
#> Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
#> 4 observations deleted due to missingness
If you find MANTA useful in your research please cite the related publication:
Garrido-Martín, D., Calvo, M., Reverter, F., Guigó, R. A fast non-parametric test of association for multiple traits. bioRxiv (2022). https://doi.org/10.1101/2022.06.06.493041