Skip to contents

Wrapper function for scTransform (v2) parameters

params_sc_sctransform.Rd

Defaults are sctransform's own with vst.flavor = "v2" applied. Only the step-1 fit scales with n_genes and n_cells: those bound the subsample the negative binomial models are fitted on, and every later pass streams gene by gene. Raising them costs fitting time, not memory.

Usage

params_sc_sctransform(
  n_genes = 2000L,
  n_cells = 2000L,
  min_cells = 5L,
  bw_adjust = 3,
  gmean_eps = 1,
  outlier_th = 10,
  poisson_diff_theta = 0.001,
  clip_min = NULL,
  clip_max = NULL
)

Arguments

n_genes

Integer. Genes in the step-1 subsample. Defaults to 2000L.

n_cells

Integer. Cells in the step-1 subsample. Defaults to 2000L.

min_cells

Integer. Minimum number of cells a gene must be detected in to be modelled. Defaults to 5L.

bw_adjust

Numeric. Bandwidth multiplier for the kernel regression that regularises the parameters. Defaults to 3.0.

gmean_eps

Numeric. Offset in the geometric mean. Defaults to 1.0.

outlier_th

Numeric. Threshold, in median absolute deviations, past which a step-1 fit is treated as an outlier. Defaults to 10.0.

poisson_diff_theta

Numeric. Below this, the fitted dispersion is taken as the Poisson limit. Defaults to 0.001.

clip_min

Numeric or NULL. Lower residual clipping bound. NULL uses -sqrt(n_cells). Must be given together with clip_max. Defaults to NULL.

clip_max

Numeric or NULL. Upper residual clipping bound. NULL uses sqrt(n_cells). Must be given together with clip_min. Defaults to NULL.

Value

A named list with the following elements:

  • n_genes - Integer. Genes in the step-1 subsample. Defaults to 2000L.

  • n_cells - Integer. Cells in the step-1 subsample. Defaults to 2000L.

  • min_cells - Integer. Minimum number of cells a gene must be detected in to be modelled. Defaults to 5L.

  • bw_adjust - Numeric. Bandwidth multiplier for the kernel regression that regularises the parameters. Defaults to 3.0.

  • gmean_eps - Numeric. Offset in the geometric mean. Defaults to 1.0.

  • outlier_th - Numeric. Threshold, in median absolute deviations, past which a step-1 fit is treated as an outlier. Defaults to 10.0.

  • poisson_diff_theta - Numeric. Below this, the fitted dispersion is taken as the Poisson limit. Defaults to 0.001.

  • clip_min - Numeric or NULL. Lower residual clipping bound. NULL uses -sqrt(n_cells). Must be given together with clip_max. Defaults to NULL.

  • clip_max - Numeric or NULL. Upper residual clipping bound. NULL uses sqrt(n_cells). Must be given together with clip_min. Defaults to NULL.

References

Choudhary and Satija, Genome Biology, 2022.