SAFE: Single-cell Aggregated clustering From Ensemble (SAFE)

In yycunc/SAFEclustering: SAFE-clustering:Single-cell Aggregated (From Ensemble) Clustering for Single-cell RNA-seq Data

Description

SAFE (Single-cell Aggregated clustering From Ensemble): Cluster ensemble for single-cell RNA-seq data

Usage

SAFE(cluster_results, program.dir = ".", k_min = 2, k_max = NULL,
  MCLA = TRUE, HGPA = FALSE, CSPA = FALSE, cspc_cell_max = NULL,
  SEED = 1)

Arguments

cluster_results	a J*N matrix with J individual clustering methods and N cells
program.dir	defines the directory of two programs shmetis and gpmetis, which are required by HGPA algorithm and MCLA and CSPA algorithms, respectively. Default is the current directory (".").
k_min	defines the minimum number of clusters used for ensembel clustering. Default is 2.
k_max	defines the maximum number of clusters used for ensembel clustering. Default is the maximum cluster number estimated by all the single solutions.
MCLA	a boolean parameter that defines whether to use MCLA algorithm for ensemble clustering. Default is "TRUE".
HGPA	a boolean parameter that defines whether to use HGPA algorithm for ensemble clustering. Default is "FALSE".
CSPA	a boolean parameter that defines whether to use CSPA algorithm for ensemble clustering. Default is "FALSE".
cspc_cell_max	defines the maximum number of cells above which CSPA is not run, when CSPA = TRUE.
SEED	sets the seed of the random number generator. Setting the seed to a fixed value can produce reproducible clustering results for MCLA and CSPA algorithms.

Value

SAFE returns a list object containing:

• optimal_clustering: optimal ensemble clustering result determined by ANMI

• optimal_k: optimal cluster number

• HGPA: optimal cluster result using HGPA algorithm

• HGPA_optimal_k: optimal cluster number estimated by HGPA algorithm

• HGPA_ANMI: Average normalized mutual information (ANMI) for the optimal cluster result using HGPA algorithm

• MCLA: optimal cluster result using MCLA algorithm

• MCLA_optimal_k: optimal cluster number estimated by MCLA algorithm

• MCLA_ANMI: ANMI for the optimal cluster result using MCLA algorithm

• CSPA: optimal cluster result using CSPA algorithm

• CSPA_optimal_k: optimal cluster number estimated by CSPA algorithm

• CSPA_ANMI: ANMI for the optimal cluster result using CSPA algorithm

• Summary: a summary of the ensemble clustering result

Author(s)

Yuchen Yang <[e-mail]>, Ruth Huh <[e-mail]>, Yun Li <[e-mail]>

References

Yuchen Yang, Ruth Huh, Houston Culpepper, Yuan Lin, Michael Love, Yun Li. SAFE (Single-cell Aggregated clustering From Ensemble): Cluster ensemble for single-cell RNA-seq data. 2017

Examples

# Load the example data data_SAFE
data("data_SAFE")
working.dir = "~/Documents/single_cell_clustering"

# Zheng dataset
# Run individual_clustering
cluster.result <- individual_clustering(inputTags=data_SAFE$Zheng.expr, SEED=123)

# Cluster ensemble using SAFE clustering:
cluster.ensemble <- SAFE(cluster_results=cluster.result, program.dir = working.dir, SEED=123)

# Biase dataset
# Run individual_clustering
cluster.result <- individual_clustering(inputTags = data_SAFE$Biase.expr, datatype = "FPKM", seurat_min_cell = 200, resolution_min = 1.2, tsne_min_cells = 200, tsne_min_perplexity = 10, SEED=123)

# Cluster ensemble using SAFE clustering:
cluster.ensemble <- SAFE(cluster_results=cluster.result, program.dir = working.dir, SEED=123)

yycunc/SAFEclustering documentation built on March 29, 2021, 5:58 a.m.