csmFinder: Find cell-subset specific methylation loci in genome.

In Gavin-Yinld/csmFinder: Find putative cell-subset specific methylation (pCSM) loci in single-cell methylomes or bulk methylomes

Description

Find cell-subset methylation segments from methylomes generated by single cells or bulk tissue. For single-cell methylomes, a beta mixture model is used for divide the cells into two subsets with hypo and hyper-methylation states in candidate CSM regions. For regular datasets generated from bulk tissue or sorted cell populations, a Nonparametric Bayesian clustering is used to group the sequencing reads into hypo and hyper-methylated subsets, both of the two algorithms identify the 4-CpG segments with biplolar methylation patterns across two subsets as the pCSM loci.

Usage

csmFinder(candidate,data_type='regular',depth=10,
          distance=0.3,pval=0.05,thread=1)

Arguments

candidate	the candidate segments used for CSM identification
data_type	"regular" and "single-cell" represents regular datasets and single-cell datasets, respectively
depth	numeric threshold represents the least number of reads (for regular datasets) or cells (for single-cell datasets) covered the candidate segments
distance	methylation difference between hypo and hyper-methylated cells subsets or reads
pval	significance of the differnece between hypo and hyper-methylated cells subsets or reads
thread	number of threads used to identify candidate pCSM segment

Value

For single-cell datasets, the output is in the same format with the output of beta mixture model (https://github.com/Evan-Evans/Beta-Mixture-Model). For regular methylomes, the output is a matrix contains the methylation difference between hypo and hyper-methylated reads, and its significance.

References

Wu, X., et al., 2015, Nonparametric Bayesian clustering to detect bipolar methylated genomic loci, BMC Bioinformatics, 16.

Luo, Y., et al., 2018, Integrative single-cell omics analyses reveal epigenetic heterogeneity in mouse embryonic stem cells, PLoS computational biology, 14, e1006034

Examples

###need object for 'candidate' from former steps.  
#for bulk methylome
#pcsm_segment <- csmFinder(candidate,data_type='regular',thread=1)

#for single-cell methylome
#scPcsm_segment <- csmFinder(scCandidate,data_type='single-cell',thread=1)

Gavin-Yinld/csmFinder documentation built on Sept. 16, 2019, 3:31 p.m.