Compute reference matrix from an `screference` object using AutoGeneS

Usage

autogenes_scref(
  scref,
  ngen = 500,
  mode = "standard",
  nfeatures = 2000,
  seed = 0,
  population_size = 100,
  offspring_size = 50,
  crossover_pb = 0.7,
  mutation_pb = 0.3,
  crossover_thresh = 1000,
  int_standard_pb = 0.1,
  ...
)

Arguments

scref: an object of class `screference`
ngen: an integer, number of generations for optimization
mode: one of `"standard"` or `"fixed"`. If "fixed", `"nfeatures"` is used.
nfeatures: integer, number of genes selected in fixed mode.
seed: int, for reproducibility
population_size: int, size of each generation (mu parameter)
offspring_size: int, number of individuals per generation (lambda parameter)
crossover_pb: float, crossover probability
mutation_pb: float, mutation probability
...: passed to [Seurat::FindVariableFeatures]
crossover_thres: int, crossover threshold (standard mode)
ind_standard_pb: float, probability used to generate initial population in standard mode

Value

an screference object updated with AutoGeneS reference centroids

Note

Reference: Aliee, Hananeh, and Fabian J. Theis. 2021. “AutoGeneS: Automatic Gene Selection Using Multi-Objective Optimization for RNA-Seq Deconvolution.” Cell Systems 12 (7): 706-715.e4. https://doi.org/10.1016/j.cels.2021.05.006.

See also: https://github.com/theislab/AutoGeneS