Main clustering CORE V2.0 updated

CORE is an algorithm to generate reproduciable clustering, CORE is first implemented in ascend R package. Here, CORE V2.0 introduces several new functionalities, including three key features: fast (and more memory efficient) implementation with C++ and paralellisation options allowing clustering of hundreds of thousands of cells (ongoing development), outlier revomal important if singletons exist (done), a number of dimensionality reduction methods including the imputation implementation (CIDR) for confirming clustering results (done), and an option to select the number of optimisation tree height windows for increasing resolution

CORE_clustering(
  mixedpop = NULL,
  windows = seq(from = 0.025, to = 1, by = 0.025),
  remove_outlier = c(0),
  nRounds = 1,
  PCA = FALSE,
  nPCs = 20,
  ngenes = 1500,
  verbose = FALSE,
  log_transform = FALSE
)

Arguments

mixedpop	is a SingleCellExperiment object from the train mixed population
windows	a numeric specifying the number of windows to test
remove_outlier	a vector containing IDs for clusters to be removed the default vector contains 0, as 0 is the cluster with singletons.
nRounds	an integer specifying the number rounds to attempt to remove outliers.
PCA	logical specifying if PCA is used before calculating distance matrix
nPCs	an integer specifying the number of principal components to use.
ngenes	number of genes used for clustering calculations.
verbose	a logical whether to display additional messages
log_transform	boolean whether log transform should be computed

Value

a list with clustering results of all iterations, and a selected optimal resolution

Author

Quan Nguyen, 2017-11-25

Examples

day5 <- day_5_cardio_cell_sample
#day5$dat5_counts needs to be in a matrix format
cellnames <- colnames(day5$dat5_counts)
cluster <-day5$dat5_clusters
cellnames <-data.frame('Cluster'=cluster, 'cellBarcodes' = cellnames)
mixedpop2 <-new_summarized_scGPS_object(ExpressionMatrix = day5$dat5_counts, 
    GeneMetadata = day5$dat5geneInfo, CellMetadata = day5$dat5_clusters)
test <- CORE_clustering(mixedpop2, remove_outlier = c(0), PCA=FALSE, nPCs=20,
    ngenes=1500)
#> Performing 1 round of filtering
#> Identifying top variable genes
#> Calculating distance matrix
#> Performing hierarchical clustering
#> Finding clustering information
#> No more outliers detected in filtering round 1
#> Identifying top variable genes
#> Calculating distance matrix
#> Performing hierarchical clustering
#> Finding clustering information
#> Done clustering, moving to stability calculation...
#> Done calculating stability...
#> Start finding optimal clustering...
#> Done finding optimal clustering...