doLeidenCluster

cluster cells using a NN-network and the Leiden community detection algorithm

doLeidenCluster(
  gobject,
  spat_unit = NULL,
  feat_type = NULL,
  name = "leiden_clus",
  nn_network_to_use = "sNN",
  network_name = "sNN.pca",
  python_path = NULL,
  resolution = 1,
  weight_col = "weight",
  partition_type = c("RBConfigurationVertexPartition", "ModularityVertexPartition"),
  init_membership = NULL,
  n_iterations = 1000,
  return_gobject = TRUE,
  set_seed = TRUE,
  seed_number = 1234
)

Arguments

gobject

giotto object

spat_unit

spatial unit (e.g. "cell")

feat_type

feature type (e.g. "rna", "dna", "protein")

name

name for cluster, default to "leiden_clus"

nn_network_to_use

type of NN network to use (kNN vs sNN), default to "sNN"

network_name

name of NN network to use, default to "sNN.pca"

python_path

specify specific path to python if required

resolution

resolution, default = 1

weight_col

weight column to use for edges, default to "weight"

partition_type

The type of partition to use for optimization. (e.g. "RBConfigurationVertexPartition", "ModularityVertexPartition")

init_membership

initial membership of cells for the partition

n_iterations

number of interactions to run the Leiden algorithm. If the number of iterations is negative, the Leiden algorithm is run until an iteration in which there was no improvement.

return_gobject

logical. return giotto object (default = TRUE)

set_seed

set seed

seed_number

number for seed

Value

giotto object with new clusters appended to cell metadata

Details

This function is a wrapper for the Leiden algorithm implemented in python, which can detect communities in graphs of millions of nodes (cells), as long as they can fit in memory. See the leidenalg github page or the readthedocs page for more information.

Partition types available and information:

• RBConfigurationVertexPartition: Implements Reichardt and Bornholdt’s Potts model with a configuration null model. This quality function is well-defined only for positive edge weights. This quality function uses a linear resolution parameter.

• ModularityVertexPartition: Implements modularity. This quality function is well-defined only for positive edge weights. It does not use the resolution parameter

Set weight_col = NULL to give equal weight (=1) to each edge.

Examples

g <- GiottoData::loadGiottoMini("visium")
#> 1. read Giotto object
#> 2. read Giotto feature information
#> 3. read Giotto spatial information
#> 3.1 read Giotto spatial shape information
#> 3.2 read Giotto spatial centroid information
#> 3.3 read Giotto spatial overlap information
#> 4. read Giotto image information
#> python already initialized in this session
#>  active environment : '/usr/bin/python3'
#>  python version : 3.10
#> checking default envname 'giotto_env'
#> a system default python environment was found
#> Using python path:
#>  "/usr/bin/python3"

doLeidenCluster(g)
#> Error in py_run_file_impl(file, local, convert): ModuleNotFoundError: No module named 'igraph'
#> Run `reticulate::py_last_error()` for details.