BASiCS_ParallelMCMC: Parallelized BASiCS_MCMC

Description Usage Arguments Details Value See Also

View source: R/BASiCS_ParallelMCMC.R

Description

Parallelize BASiCS normalization implemented in BASiCS_MCMC.

Usage

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
BASiCS_ParallelMCMC(
  CountsList,
  SpikeInfo,
  N = 200000L,
  Burn = 80000L,
  Thin = 40L,
  Regression = TRUE,
  MinCount = 10L,
  PropMinCount = 0.85,
  NCores = 2L,
  Name = paste0(sample(c(0:9, letters), 10), collapse = ""),
  Verbose = TRUE
)

Arguments

CountsList

A list of unfiltered count matrices with genes as rows and samples as columns.

SpikeInfo

data.frame whose first and second columns contain the gene names assigned to the spike-in genes (they must match the ones in rownames(Counts)) and the associated input number of molecules, respectively. If SpikeInfo = NULL, only the horizontal integration implementation (no spikes) can be run. Default value: SpikeInfo = NULL.

N

Total number of iterations for the MCMC sampler. Use N>=max(4,Thin), N being a multiple of Thin.

Burn

Burn-in period for the MCMC sampler. Use Burn>=1, Burn<N, Burn being a multiple of Thin.

Thin

Thining period for the MCMC sampler. Use Thin>=2.

Regression

If Regression = TRUE, BASiCS exploits a joint prior formulation for mean and over-dispersion parameters to estimate a measure of residual over-dispersion is not confounded by mean expression. Recommended setting is Regression = TRUE.

MinCount

integer; Minimum number of counts for a gene.

Name

character; Chain identification name.

PropMinCount;

numeric; Proportion of samples with minimum count.

NCores;

integer; Number of cores.

Verbose;

bool; Print progress in ./logs directory?

Details

Assumes that there are spike-ins and that these are named "ERCC". Filter rule: a gene must be counted at least MinCount in PropMinCount percent of the samples. For biological transcripts, this needs to be true in at least one group. For spike-ins in all the groups. Uses mclapply to parallelize.

Value

A list of BASiCS_Chain object.

See Also

BASiCS_Chain

BASiCS_MCMC


oliviermfmartin/HelpingHand documentation built on Oct. 10, 2020, 5:59 a.m.