ziaq_fit: Zero-inflation adjusted quantile regression

In gefeizhang/ZIAQ: A quantile regression method for differential expression analysis of single-cell RNA-seq data

Description

This function fits the zero-inflation adjusted quantile regression model for one individual gene.

Usage

ziaq_fit(
  y,
  colDat,
  formula = ~condition,
  group = "condition",
  probs = c(0.25, 0.5, 0.75),
  log_i = T
)

Arguments

y	a numeric vector including gene expression value for one gen
colDat	a dataframe including the individual cell information
formula	a formula with the predictors included in colDat. The default is ~condition.
group	the variable name in colDat for the factor used in group comparsion. he default is condition.
probs	the quantile levels for the quantile regressions. The default is c(0.25, 0.5, 0.75)
log_i	TRUE or FALSE indicate whether to apply log transformation. The default is TRUE.

Value

pvalue	The p-value for testing the signficance of the specified group variable.
logistic_model	The glm object of the logistics regression for fitting binary outcome (zero or not zero) vs the predictors specified in formula.
zero_proportion	The proportion of zero measurements
quantile_levels	The input quantile levels and zero-proportion adjusted quantile levels
qr	The objects of the fitted quantile regressions

Examples

# simulated data
y = round(100* runif(100))
colDat = data.frame(condition = rep(c(1, 0), e = 50))
res = ziaq_fit(y, colDat = colDat,  formula = ~ condition,
              group = 'condition', probs = c(0.25, 0.5, 0.75),
              log_i = TRUE)
print(res)

gefeizhang/ZIAQ documentation built on Feb. 26, 2020, 6:15 p.m.