splinectomeR

R package of spline-based statistical analysis tools for longitudinal data

These functions are designed to provide statistical analyses in real longitudinal data, which may be missing timepoints, possess limited data at some timepoints, have noisy biological variability, and variable numbers of observations per individual being measured. Comparisons can be made between two groups or within a single group for a non-zero change over the independent axis; both return a p-value based on a randomly permuted distribution of the real data. There is also a function for measuring significant differences at intervals across the entire x series (e.g. time) by interpolating splines from the original data.

For information on installation of the package from GitHub, and a basic outline of the functions, go here.

Happy splining, and have a splinedid day!

Vignettes

There are two vignettes available for this package:

1. Chick Weights - a simple exploration and splinectomeR tutorial using the R dataset ChickWeight.
2. Yassour et al. Antibiotics - a more intensive application of the splinectomeR package to a large set of longitudinal microbiome data published by Yassour et al.¹ following 38 babies over three years.

Disclaimer

Licensed under the MIT License. As such, the splinectomer package is freely offered, without any warrantee or guarantee. It may be used, redistributed, or modified for non-commercial purposes with appropriate citation of the original work. You, the user, are solely responsible for verifying the legitimacy of your results.

Citation

Shields-Cutler RR, Al-Ghalith GA, Yassour M, Knights D. (2018) SplinectomeR Enables Group Comparisons in Longitudinal Microbiome Studies. Frontiers in Microbiology 9:785. doi: 10.3389/fmicb.2018.00785

References