Performs BBKNN integration — perform.bbknn • IBRAP

Performs BBKNN integration on defined method-assays and reductions contained within. This is performed on reductions.

perform.bbknn(
  object,
  assay,
  reduction,
  graph.name.suffix = "",
  batch,
  approx = FALSE,
  metric = "euclidean",
  neighbors_within_batch = 3,
  n_pcs = NULL,
  trim = NULL,
  annoy_n_trees = 10,
  use_faiss = TRUE,
  set_op_mix_ratio = 1,
  local_connectivity = 1,
  generate.diffmap = FALSE,
  n_comps = 15,
  diffmap.name.suffix = "",
  verbose = FALSE,
  seed = 1234
)

Arguments

object: IBRAP S4 class object
assay: Character. String containing indicating which assay to use
reduction: Character. String defining the name of the reduction to provide for BBKNN. Default = NULL
graph.name.suffix: Character. Should a suffix be added to the end of bbknn as the graph name, i.e. parameter changes?
batch: Character. Column name in metadata indicating batch. Can be multiple.
approx: Character. Employs annoy's approximate neighbour finding. Useful for large datasets but may increase correction.
neighbors_within_batch: Numerical. How many neighbours to report per batch. Default = 3
n_pcs: Numerical. Range of principal components to use. Default = NULL
trim: Numerical. Trims the n of neighbours per cell to this value. Helps with population independence. Default = NULL
annoy_n_trees: Numerical. Number of trees to generate in annoy forest. More trees provides higher precision at the cost of increased resource demand and run time. Default = 10
use_faiss: Boolean. Uses faiss package to compute nearest neighbour, this improves run time at the cost of precision. Default = TRUE
set_op_mix_ratio: Numerical. UMAP connectivity parameter between 0 and 1. controls the blen d between a connectivity matrix formed exclusively from mutual nearest neighbour pairs (0) and a union of all observed neighbour relationships with the mutual pairs emphasised (1). Default = 1.0
local_connectivity: Numerical. How many nearest neighbours of each cell are assumed to be fully connected. Default = 1
generate.diffmap: Boolean. Should diffusion maps be generated from the neighourhood graphs, these will be stored in computational_reductions and can be used for umap generation and further neighbourhood generation. Default = TRUE
n_comps: Numerical. How many components should be generated for the diffusion maps. Default = 15
verbose: Logical Should function messages be printed?
seed: Numerical What seed should be set. Default = 1234
metric.: Character. Which distance metric to use when approx is TRUE, options: 'angular', 'euclidean', 'manhattan' or 'hamming'. Default = 'euclidean'
diffmap.name.sufix: Character. Should a suffix be added to the end of bbknn:diffmap as the reduction name, i.e. parameter changes?

Value

BBKNN connectivity graph contained in graphs in the indicated method-assays

Examples


object <- perform.bbknn(object = object, 
                        assay = c('SCT', 'SCANPY', 'SCRAN'), 
                        reduction = c('pca'),
                        batch = 'tech')
#> Error in is(object = object, class2 = "IBRAP"): object 'object' not found