SnapATAC_BinarizeBmat: Binarize Bmat according SnapATAC

In beyondpie/smmtools: Single-cell multi-omics sequenceing data tools

Binarize Bmat according SnapATAC

Description

• Filter features if the covariances are smaller or larger than thresholds.

• Too high values will be set as zeros.

Usage

SnapATAC_BinarizeBmat(
  bmat,
  cell_by_feature = FALSE,
  mu = NULL,
  sigma = NULL,
  z_threshold = 1.65,
  outlier = 0.001
)

Arguments

bmat	sparse matrix, feature by cells (default) or verse In SnapATAC package, bmat is cell by feature. Be sure to claim the format with the parameter cell_by_feature below.
cell_by_feature,	bool, if bmat is cell by feature or not, default is FALSE
mu	double or NULL, mean of log10 count to get z score, default is NULL, which will be estimated from the data.
sigma	double or NULL, standard deviation of log10 count to get z score, default is NULL, which will be estimated from the data.
z_threshold	double, z score threshold, features highter than it or lower than -it, will be removed, default is 1.65
outlier	double, count larger than quantile of 1- outlier will be treated as zero, default is 1e-3.

Value

list of two elements

• bmat sparse matrix, cell by feature (consistent with SnapATAC)

• featureIndexKept integer vector, feature index kept after filtering

• mu double, mean for z_score

• sigma double, standard deviation for z_score

• z_threshold double, cutoff for z_threshold

• outlier double, cutoff for count

beyondpie/smmtools documentation built on July 1, 2022, 4:33 a.m.