SplatPopParams: The SplatPopParams class

In Oshlack/splatter: Simple Simulation of Single-cell RNA Sequencing Data

The SplatPopParams class

Description

S4 class that holds parameters for the splatPop simulation.

Parameters

In addition to the SplatParams parameters, splatPop simulation requires the following parameters:

[similarity.scale]

Scaling factor for pop.cv.param.rate, where values larger than 1 increase the similarity between individuals in the population and values less than one make the individuals less similar.

[eqtl.n]

The number (>1) or percent (<=1) of genes to assign eQTL effects.

[eqtl.dist]

Maximum distance between eSNP and eGene

[eqtl.maf.min]

Minimum Minor Allele Frequency of eSNPs.

[eqtl.maf.max]

Maximum Minor Allele Frequency of eSNPs.

[eqtl.coreg]

Proportion of eGenes to have a shared eSNP (i.e., co-regulated genes)

[eqtl.group.specific]

Percent of eQTL effects to simulate as group specific.

[eqtl.condition.specific]

Percent of eQTL effects to simulate as condition specific.

eQTL Effect size distribution parameters. Defaults estimated from GTEx eQTL mapping results, see vignette for more information.

eqtl.ES.shape

Shape parameter for the effect size gamma distribution.

eqtl.ES.rate

Rate parameter for the effect size gamma distribution.

Bulk Mean Expression distribution parameters. Defaults estimated from GTEx data, see vignette for more information.

pop.mean.shape

Shape parameter for the mean (i.e. bulk) expression gamma distribution

pop.mean.rate

Rate parameter for the mean (i.e. bulk) expression gamma distribution

Bulk Expression Coefficient of Variation distribution parameters binned. Defaults estimated from GTEx data, see vignette for more information.

pop.cv.param

Dataframe containing gene mean bin range, and the CV shape, and CV rate parameters for each of those bins.

Specify number of samples per batch. Note that splatPop will randomly assign donors to be present in multiple batches to fulfill the specified nBatches and batch.size parameters. For example, if 10 samples are simulated with batchPool.n=4 and batchPool.size= 4, then 6 samples will be randomly chosen to be replicated in two pools.

batch.size

The number of donors in each pool/batch.

Specify shape and rate of gamma distribution to sample number of cells per batch per donor. Will only be used if nCells parameter is set to 0.

nCells.sample

True/False if nCells should be set as nCells or sampled from a gamma distribution for each batch/donor.

nCells.shape

Shape parameter for the nCells per batch per donor distribution.

nCells.rate

Rate parameter for the nCells per batch per donor distribution.

Condition/treatment differential expression parameters

[nConditions]

The number of conditions/treatments to divide samples into.

[condition.prob]

Probability that a sample belongs to each condition/treatment group. Can be a vector.

[cde.prob]

Probability that a gene is differentially expressed in a condition group. Can be a vector.

[cde.downProb]

Probability that a conditionally differentially expressed gene is down-regulated. Can be a vector.

[cde.facLoc]

Location (meanlog) parameter for the conditional differential expression factor log-normal distribution. Can be a vector.

[cde.facScale]

Scale (sdlog) parameter for the conditional differential expression factor log-normal distribution. Can be a vector.

The parameters not shown in brackets can be estimated from real data using splatPopEstimate. For details of the eQTL simulation see splatPopSimulate.

Oshlack/splatter documentation built on Dec. 10, 2024, 3:48 p.m.