knn_refine

Assigns the label that most neighbors have, given it is more than
<code>min_knn_prob</code>.
I've found empirically on the MALT data that <code>min_knn_prob=0.5</code> gives
good results, whether you classify the entire data set or just a single
cell type.
It simply excludes some of the cells that have more than 2 cell types in their
neighborhood and none is much stronger than the others,
so this is a reasonable, conservative filtering.

Annotate single-cell RNA sequencing data manually based on
marker gene thresholds.
Find cell type rules (gene+threshold) through exploration,
use the popular piping operator '%>%' to reconstruct complex
cell type hierarchies.
'cellpypes' models technical noise to find positive and negative cells for
a given expression threshold and returns cell type labels or pseudobulks.
Cite this package as Frauhammer (2022) <doi:10.5281/zenodo.6555728> and
visit <https://github.com/FelixTheStudent/cellpypes> for tutorials and newest
features.

Felix Frauhammer

cellpypes

Cell Type Pipes for Single-Cell RNA Sequencing Data

knn_refine function

<dl><dt>labels</dt>
<dd>Cell type labels as character or factor.</dd>
<dt>neighbors</dt>
<dd>Neighbor graph, pass <code>obj$neighbors</code>.</dd>
<dt>min_knn_prob</dt>
<dd>Value between 0 and 1, defaults to 0.5.
If the 'winning label' is below this proportion of kNN that have it,
knn_refine will return "Unassigned".</dd></dl>

Arguments

Refine cell type labels with knn classifier — knn_refine

<dl>

<dt>labels</dt>
<dd>Cell type labels as character or factor.</dd>


<dt>neighbors</dt>
<dd>Neighbor graph, pass <code>obj$neighbors</code>.</dd>


<dt>min_knn_prob</dt>
<dd>Value between 0 and 1, defaults to 0.5.
If the 'winning label' is below this proportion of kNN that have it,
knn_refine will return "Unassigned".</dd>

</dl>

knn_refine: Refine cell type labels with knn classifier

Description

Usage

Value

Arguments