spatialNetSmooth is an R package for smoothing of GSEA-scores from spatial omics breast cancer data. The smoothing is based on network-propagation and uses the netSmooth package. It contains functions that work with different methods based on spatial and/or profile-neighborhood. The input can be either a Seurat or a VoltRon object.
- Calculating GSEA-scores using gene lists from ikarus or custom list
- spatial smoothing
- NN/SNN smoothing
- spatial+NN/SNN smoothing
- union/intersection/linear combination of NN/SNN and spatial smoothing
- Calculating F1-scores of different thresholds based on quantiles
- Calculating best thresholds based on ROC-curve
- Spatial plotting of prediction compared to known ground truth (thresholded with ROC-curve)
if (!requireNamespace("BiocManager", quietly=TRUE))
install.packages("BiocManager")
if (!requireNamespace("devtools", quietly=TRUE))
install.packages("devtools")
install_github("BIMSBbioinfo/spatialNetSmooth", build_vignettes=FALSE,
dependencies=TRUE)For using VoltRon objects, VoltRon has to be installed. See the instructions for installation.
See the Tutorial