kde2d_faster

Based on the MASS kde2d() function, but heavily simplified; it's just tcrossprod() now.

One key exploratory analysis step in single-cell genomics data analysis
is the prediction of features with different activity levels. For example, we want
to predict differentially expressed genes (DEGs) in single-cell RNA-seq data,
spatial DEGs in spatial transcriptomics data, or differentially accessible
regions (DARs) in single-cell ATAC-seq data. 'singleCellHaystack' predicts differentially
active features in single cell omics datasets without relying on the clustering
of cells into arbitrary clusters. 'singleCellHaystack' uses Kullback-Leibler
divergence to find features (e.g., genes, genomic regions, etc) that are active
in subsets of cells that are non-randomly positioned inside an input space (such as
1D trajectories, 2D tissue sections, multi-dimensional embeddings, etc). For
the theoretical background of 'singleCellHaystack' we refer to our original paper
Vandenbon and Diez (Nature Communications, 2020) <doi:10.1038/s41467-020-17900-3>
and our update Vandenbon and Diez (Scientific Reports, 2023) <doi:10.1038/s41598-023-38965-2>.

Alexis Vandenbon

singleCellHaystack

A Universal Differential Expression Prediction Tool for
Single-Cell and Spatial Genomics Data

Diego Diez

kde2d_faster function

<dl><dt>dens.x</dt>
<dd>Contribution of all cells to densities of the x-axis grid points.</dd>
<dt>dens.y</dt>
<dd>Contribution of all cells to densities of the y-axis grid points.</dd></dl>

Arguments

Based on the MASS kde2d() function, but heavily simplified; it's just tcrossprod() now. — kde2d_faster

<dl>

<dt>dens.x</dt>
<dd>Contribution of all cells to densities of the x-axis grid points.</dd>


<dt>dens.y</dt>
<dd>Contribution of all cells to densities of the y-axis grid points.</dd>

</dl>

kde2d_faster: Based on the MASS kde2d() function, but heavily simplified; it's just tcrossprod() now.

Description

Usage

Arguments