plotPP

This function plots various statistics for the posterior check

Application of 'RaceID' allows inference of cell types and prediction of lineage trees by the 'StemID2' algorithm (Herman, J.S., Sagar, Grun D. (2018) <DOI:10.1038/nmeth.4662>). 'VarID2' is part of this package and allows quantification of biological gene expression noise at single-cell resolution (Rosales-Alvarez, R.E., Rettkowski, J., Herman, J.S., Dumbovic, G., Cabezas-Wallscheid, N., Grun, D. (2023) <DOI:10.1186/s13059-023-02974-1>).

Dominic Grc3<bc>n

RaceID

Identification of Cell Types, Inference of Lineage Trees, and
Prediction of Noise Dynamics from Single-Cell RNA-Seq Data

Dominic Grün

plotPP function

<dl><dt>pp</dt>
<dd>List object returned by <code>testPrior</code> function.</dd>
<dt>y</dt>
<dd>One of "mean", "median", "var", "cor", or <code>NULL</code>. If <code>NULL</code> then the ratios between the predicted and the actual variances across all sampled genes and neighbourhoods are shown as boxplots for all tested values of the prior parameter <code>gamma</code>. If <code>y</code> equals "mean", "median", or "var", the mean, median, or variance is plotted for all <code>gamma</code> values. If <code>y</code> equal "cor", then the correlation between the total transcript count of a cell and the local noise estimate <code>epsilon</code> is plotted for all values of <code>gamma</code>. Default is <code>NULL</code>.</dd>
<dt>umi.eps</dt>
<dd>Logical. If <code>TRUE</code> then a scatter plot of the local noise estimate <code>epsilon</code> and the total transcript count is produced for a given element <code>i</code> of the <code>pp$noise</code> corresponding to a value of the prior parameter <code>gamma</code>. Default is <code>FALSE</code>.</dd>
<dt>i</dt>
<dd>Positive integer number. Index of <code>pp$noise</code>, corresponding to a value of the prior parameter <code>gamma</code> to be used for plotting is <code>umi.eps=TRUE</code>. Default is 1.</dd>
<dt>log.scale</dt>
<dd>Logical. If <code>TRUE</code> then the ratio between the predicted and the actual variance is transformed to a log2-scale prior to computations and plotting. If <code>umi.eps=TRUE</code>, total transcript counts and <code>epsilon</code> estimates are log2-transformed for plotting. Default is <code>TRUE</code>.</dd></dl>

Arguments

Plotting function for posterior checks — plotPP

<dl>

<dt>pp</dt>
<dd>List object returned by <code>testPrior</code> function.</dd>


<dt>y</dt>
<dd>One of "mean", "median", "var", "cor", or <code>NULL</code>. If <code>NULL</code> then the ratios between the predicted and the actual variances across all sampled genes and neighbourhoods are shown as boxplots for all tested values of the prior parameter <code>gamma</code>. If <code>y</code> equals "mean", "median", or "var", the mean, median, or variance is plotted for all <code>gamma</code> values. If <code>y</code> equal "cor", then the correlation between the total transcript count of a cell and the local noise estimate <code>epsilon</code> is plotted for all values of <code>gamma</code>. Default is <code>NULL</code>.</dd>


<dt>umi.eps</dt>
<dd>Logical. If <code>TRUE</code> then a scatter plot of the local noise estimate <code>epsilon</code> and the total transcript count is produced for a given element <code>i</code> of the <code>pp$noise</code> corresponding to a value of the prior parameter <code>gamma</code>. Default is <code>FALSE</code>.</dd>


<dt>i</dt>
<dd>Positive integer number. Index of <code>pp$noise</code>, corresponding to a value of the prior parameter <code>gamma</code> to be used for plotting is <code>umi.eps=TRUE</code>. Default is 1.</dd>


<dt>log.scale</dt>
<dd>Logical. If <code>TRUE</code> then the ratio between the predicted and the actual variance is transformed to a log2-scale prior to computations and plotting. If <code>umi.eps=TRUE</code>, total transcript counts and <code>epsilon</code> estimates are log2-transformed for plotting. Default is <code>TRUE</code>.</dd>

</dl>

plotPP: Plotting function for posterior checks

Description

Usage

Arguments