diff --git a/R/AllClasses.R b/R/AllClasses.R
index c8066bd05d9353b0106d3e5521c3fa58afcac6c5..568b391450cd5124b83af544bf93130f2c20ea1c 100644
--- a/R/AllClasses.R
+++ b/R/AllClasses.R
@@ -96,8 +96,6 @@ setClass("SimpleParams",
#' \describe{
#' \item{\code{out.prob}}{Probability that a gene is an expression
#' outlier.}
-#' \item{\code{out.loProb}}{Probability that an expression outlier
-#' gene is lowly expressed.}
#' \item{\code{out.facLoc}}{Location (meanlog) parameter for the
#' expression outlier factor log-normal distribution.}
#' \item{\code{out.facScale}}{Scale (sdlog) parameter for the
@@ -182,7 +180,6 @@ setClass("SplatParams",
lib.loc = "numeric",
lib.scale = "numeric",
out.prob = "numeric",
- out.loProb = "numeric",
out.facLoc = "numeric",
out.facScale = "numeric",
de.prob = "numeric",
@@ -202,20 +199,19 @@ setClass("SplatParams",
prototype = prototype(nGroups = 1,
groupCells = 100,
mean.rate = 0.3,
- mean.shape = 0.4,
- lib.loc = 10,
- lib.scale = 0.5,
- out.prob = 0.1,
- out.loProb = 0.5,
+ mean.shape = 0.6,
+ lib.loc = 11,
+ lib.scale = 0.2,
+ out.prob = 0.05,
out.facLoc = 4,
- out.facScale = 1,
+ out.facScale = 0.5,
de.prob = 0.1,
de.downProb = 0.5,
de.facLoc = 4,
de.facScale = 1,
bcv.common = 0.1,
- bcv.df = 25,
- dropout.present = TRUE,
+ bcv.df = 60,
+ dropout.present = FALSE,
dropout.mid = 0,
dropout.shape = -1,
path.from = 0,
@@ -438,4 +434,4 @@ setClass("SCDDParams",
sd.range = c(1, 3),
modeFC = c(2, 3, 4),
varInflation = c(1, 1),
- condition = "condition"))
\ No newline at end of file
+ condition = "condition"))
diff --git a/R/SplatParams-methods.R b/R/SplatParams-methods.R
index e3f773d6c26622dfd475f57110eb50c608019b43..07f9916d2208b24a478d3d3155380b7794a5186e 100644
--- a/R/SplatParams-methods.R
+++ b/R/SplatParams-methods.R
@@ -27,11 +27,6 @@ setValidity("SplatParams", function(object) {
lib.loc = checkNumber(v$lib.loc),
lib.scale = checkNumber(v$lib.scale, lower = 0),
out.prob = checkNumber(v$out.prob, lower = 0, upper = 1),
- out.loProb = checkNumber(v$out.loProb, lower = 0, upper = 1),
- out.facLoc = checkNumber(v$lib.loc),
- out.facScale = checkNumber(v$lib.scale, lower = 0),
- out.prob = checkNumber(v$out.prob, lower = 0, upper = 1),
- out.loProb = checkNumber(v$out.loProb, lower = 0, upper = 1),
out.facLoc = checkNumber(v$out.facLoc),
out.facScale = checkNumber(v$out.facScale, lower = 0),
de.prob = checkNumeric(v$de.prob, lower = 0, upper = 1,
@@ -108,7 +103,6 @@ setMethod("show", "SplatParams", function(object) {
"Library size:" = c("(Location)" = "lib.loc",
"(Scale)" = "lib.scale"),
"Exprs outliers:" = c("(Probability)" = "out.prob",
- "(Lo Prob)" = "out.loProb",
"(Location)" = "out.facLoc",
"(Scale)" = "out.facScale"),
"Diff expr:" = c("[Probability]" = "de.prob",
diff --git a/R/splat-estimate.R b/R/splat-estimate.R
index 69ddec5e3b500c46fc8a8b75d698db7135c7d60b..d7a070e660e868e8f5e8001b7b2704a739797dc9 100644
--- a/R/splat-estimate.R
+++ b/R/splat-estimate.R
@@ -59,19 +59,34 @@ splatEstimate.matrix <- function(counts, params = newSplatParams()) {
#' Estimate Splat mean parameters
#'
#' Estimate rate and shape parameters for the gamma distribution used to
-#' simulate gene expression means using the 'moment matching estimation' method
-#' of \code{\link[fitdistrplus]{fitdist}}.
+#' simulate gene expression means.
#'
#' @param norm.counts library size normalised counts matrix.
#' @param params SplatParams object to store estimated values in.
#'
+#' @details
+#' Parameter for the gamma distribution are estimated by fitting the mean
+#' normalised counts using \code{\link[fitdistrplus]{fitdist}}. The 'maximum
+#' goodness-of-fit estimation' method is used to minimise the Cramer-von Mises
+#' distance. This can fail in some situations, in which case the 'method of
+#' moments estimation' method is used instead. Prior to fitting the means are
+#' winsorized by setting the top and bottom 10 percent of values to the 10th
+#' and 90th percentiles.
+#'
#' @return SplatParams object with estimated values.
splatEstMean <- function(norm.counts, params) {
means <- rowMeans(norm.counts)
means <- means[means != 0]
- fit <- fitdistrplus::fitdist(means, "gamma", method = "mme")
+ means <- winsorize(means, q = 0.1)
+
+ fit <- try(fitdistrplus::fitdist(means, "gamma", method = "mge",
+ gof = "CvM"))
+ if (class(fit) == "try-error") {
+ warning("Goodness of fit failed, using Method of Moments")
+ fit <- fitdistrplus::fitdist(means, "gamma", method = "mme")
+ }
params <- setParams(params, mean.shape = unname(fit$estimate["shape"]),
mean.rate = unname(fit$estimate["rate"]))
@@ -111,15 +126,12 @@ splatEstLib <- function(counts, params) {
#' @details
#' Expression outlier genes are detected using the Median Absolute Deviation
#' (MAD) from median method. If the log2 mean expression of a gene is greater
-#' than two MADs from the median log2 mean expression it is designated as a
+#' than two MADs above the median log2 mean expression it is designated as an
#' outlier. The proportion of outlier genes is used to estimate the outlier
-#' probability. The low outlier probability is estimated as the proportion of
-#' outlier genes that have a log2 mean less than the median log2 mean. Factors
-#' for each outlier gene are calculated by dividing mean expression by the
-#' median mean expression. A log-normal distribution is then fitted to these
-#' factors in order to estimate the outlier factor location and scale
-#' parameters. See \code{\link[fitdistrplus]{fitdist}} for details on the
-#' fitting.
+#' probability. Factors for each outlier gene are calculated by dividing mean
+#' expression by the median mean expression. A log-normal distribution is then
+#' fitted to these factors in order to estimate the outlier factor location and
+#' scale parameters using \code{\link[fitdistrplus]{fitdist}}.
#'
#' @return SplatParams object with estimated values.
splatEstOutlier <- function(norm.counts, params) {
@@ -130,34 +142,42 @@ splatEstOutlier <- function(norm.counts, params) {
med <- median(lmeans)
mad <- mad(lmeans)
- lo.bound <- med - 2 * mad
- hi.bound <- med + 2 * mad
+ bound <- med + 2 * mad
+
+ outs <- which(lmeans > bound)
- lo.outs <- which(lmeans < lo.bound)
- hi.outs <- which(lmeans > hi.bound)
+ prob <- length(outs) / nrow(norm.counts)
- prob <- (length(lo.outs) + length(hi.outs)) / nrow(norm.counts)
- lo.prob <- length(lo.outs) / (length(lo.outs) + length(hi.outs))
+ params <- setParams(params, out.prob = prob)
- facs <- means[c(lo.outs, hi.outs)] / median(means)
- fit <- fitdistrplus::fitdist(facs, "lnorm")
+ if (length(outs) > 1) {
+ facs <- means[outs] / median(means)
+ fit <- fitdistrplus::fitdist(facs, "lnorm")
- params <- setParams(params, out.prob = prob, out.loProb = lo.prob,
- out.facLoc = unname(fit$estimate["meanlog"]),
- out.facScale = unname(fit$estimate["sdlog"]))
+ params <- setParams(params,
+ out.facLoc = unname(fit$estimate["meanlog"]),
+ out.facScale = unname(fit$estimate["sdlog"]))
+ }
return(params)
}
#' Estimate Splat Biological Coefficient of Variation parameters
#'
-#' Parameters are estimated using the \code{estimateDisp} function in the
-#' \code{edgeR} package. Specifically the common dispersion and prior degrees
-#' of freedom. See \code{\link{estimateDisp}} for details.
+#' Parameters are estimated using the \code{\link[edgeR]{estimateDisp}} function
+#' in the \code{edgeR} package.
#'
#' @param counts counts matrix to estimate parameters from.
#' @param params SplatParams object to store estimated values in.
#'
+#' @details
+#' The \code{\link[edgeR]{estimateDisp}} function is used to estimate the common
+#' dispersion and prior degrees of freedom. See
+#' \code{\link[edgeR]{estimateDisp}} for details. When estimating parameters on
+#' simulated data we found a broadly linear relationship between the true
+#' underlying common dispersion and the \code{edgR} estimate, therefore we
+#' apply a small correction, \code{disp = 0.1 + 0.25 * edgeR.disp}.
+#'
#' @return SplatParams object with estimated values.
splatEstBCV <- function(counts, params) {
@@ -165,7 +185,8 @@ splatEstBCV <- function(counts, params) {
design <- matrix(1, ncol(counts), 1)
disps <- edgeR::estimateDisp(counts, design = design)
- params <- setParams(params, bcv.common = disps$common.dispersion,
+ params <- setParams(params,
+ bcv.common = 0.1 + 0.25 * disps$common.dispersion,
bcv.df = disps$prior.df)
return(params)
@@ -224,4 +245,4 @@ splatEstDropout <- function(norm.counts, params) {
dropout.shape = shape)
return(params)
-}
\ No newline at end of file
+}
diff --git a/R/splat-simulate.R b/R/splat-simulate.R
index 4e3efb0df119ceaf49aa566a37d355e3fdf32c74..bb54cb6f870cfa097a5761a04ad382d681c960a4 100644
--- a/R/splat-simulate.R
+++ b/R/splat-simulate.R
@@ -280,7 +280,6 @@ splatSimGeneMeans <- function(sim, params) {
mean.shape <- getParam(params, "mean.shape")
mean.rate <- getParam(params, "mean.rate")
out.prob <- getParam(params, "out.prob")
- out.loProb <- getParam(params, "out.loProb")
out.facLoc <- getParam(params, "out.facLoc")
out.facScale <- getParam(params, "out.facScale")
@@ -288,7 +287,7 @@ splatSimGeneMeans <- function(sim, params) {
base.means.gene <- rgamma(nGenes, shape = mean.shape, rate = mean.rate)
# Add expression outliers
- outlier.facs <- getLNormFactors(nGenes, out.prob, out.loProb, out.facLoc,
+ outlier.facs <- getLNormFactors(nGenes, out.prob, 0, out.facLoc,
out.facScale)
median.means.gene <- median(base.means.gene)
outlier.means <- median.means.gene * outlier.facs
diff --git a/R/utils.R b/R/utils.R
index 73bb02308363de31fd84e839968846a77d29bca7..39e770d8932c8bb185df020cbd45fc54bdf5a47a 100644
--- a/R/utils.R
+++ b/R/utils.R
@@ -26,4 +26,27 @@ rbindMatched <- function(df1, df2) {
combined <- rbind(df1[, common.names], df2[, common.names])
return(combined)
-}
\ No newline at end of file
+}
+
+#' Winsorize vector
+#'
+#' Set outliers in a numeric vector to a specified percentile.
+#'
+#' @param x Numeric vector to winsorize
+#' @param q Percentile to set from each end
+#'
+#' @return Winsorized numeric vector
+winsorize <- function(x, q) {
+
+ checkmate::check_numeric(x, any.missing = FALSE)
+ checkmate::check_number(q, lower = 0, upper = 1)
+
+ lohi <- stats::quantile(x, c(q, 1 - q), na.rm = TRUE)
+
+ if (diff(lohi) < 0) { lohi <- rev(lohi) }
+
+ x[!is.na(x) & x < lohi[1]] <- lohi[1]
+ x[!is.na(x) & x > lohi[2]] <- lohi[2]
+
+ return(x)
+}
diff --git a/man/SplatParams.Rd b/man/SplatParams.Rd
index 473507c2e3edc05f9886a1a2af66369e59c2fe69..14bca2a036705ada0b92b2e5084609ed98075a48 100644
--- a/man/SplatParams.Rd
+++ b/man/SplatParams.Rd
@@ -40,8 +40,6 @@ The Splatter simulation requires the following parameters:
\describe{
\item{\code{out.prob}}{Probability that a gene is an expression
outlier.}
- \item{\code{out.loProb}}{Probability that an expression outlier
- gene is lowly expressed.}
\item{\code{out.facLoc}}{Location (meanlog) parameter for the
expression outlier factor log-normal distribution.}
\item{\code{out.facScale}}{Scale (sdlog) parameter for the
diff --git a/man/splatEstBCV.Rd b/man/splatEstBCV.Rd
index 9b66b1706a9ec2e31aaae4515c704c1a34bc668d..5f81642f3d4f4252c865fd31713520bc3e6fca22 100644
--- a/man/splatEstBCV.Rd
+++ b/man/splatEstBCV.Rd
@@ -15,7 +15,14 @@ splatEstBCV(counts, params)
SplatParams object with estimated values.
}
\description{
-Parameters are estimated using the \code{estimateDisp} function in the
-\code{edgeR} package. Specifically the common dispersion and prior degrees
-of freedom. See \code{\link{estimateDisp}} for details.
+Parameters are estimated using the \code{\link[edgeR]{estimateDisp}} function
+in the \code{edgeR} package.
+}
+\details{
+The \code{\link[edgeR]{estimateDisp}} function is used to estimate the common
+dispersion and prior degrees of freedom. See
+\code{\link[edgeR]{estimateDisp}} for details. When estimating parameters on
+simulated data we found a broadly linear relationship between the true
+underlying common dispersion and the \code{edgR} estimate, therefore we
+apply a small correction, \code{disp = 0.1 + 0.25 * edgeR.disp}.
}
diff --git a/man/splatEstMean.Rd b/man/splatEstMean.Rd
index e0892f9308585edab292d8eac1d38dbb1d5be812..c019f6782a4bb61e4ca5605b09d65ad9725a07e9 100644
--- a/man/splatEstMean.Rd
+++ b/man/splatEstMean.Rd
@@ -16,6 +16,14 @@ SplatParams object with estimated values.
}
\description{
Estimate rate and shape parameters for the gamma distribution used to
-simulate gene expression means using the 'moment matching estimation' method
-of \code{\link[fitdistrplus]{fitdist}}.
+simulate gene expression means.
+}
+\details{
+Parameter for the gamma distribution are estimated by fitting the mean
+normalised counts using \code{\link[fitdistrplus]{fitdist}}. The 'maximum
+goodness-of-fit estimation' method is used to minimise the Cramer-von Mises
+distance. This can fail in some situations, in which case the 'method of
+moments estimation' method is used instead. Prior to fitting the means are
+winsorized by setting the top and bottom 10 percent of values to the 10th
+and 90th percentiles.
}
diff --git a/man/splatEstOutlier.Rd b/man/splatEstOutlier.Rd
index 81e56853061557d9b4618004bd8a570a53eb0575..543b6e6337c095b81af6053fa6374d4c89172fa4 100644
--- a/man/splatEstOutlier.Rd
+++ b/man/splatEstOutlier.Rd
@@ -21,13 +21,10 @@ median mean expression level.
\details{
Expression outlier genes are detected using the Median Absolute Deviation
(MAD) from median method. If the log2 mean expression of a gene is greater
-than two MADs from the median log2 mean expression it is designated as a
+than two MADs above the median log2 mean expression it is designated as an
outlier. The proportion of outlier genes is used to estimate the outlier
-probability. The low outlier probability is estimated as the proportion of
-outlier genes that have a log2 mean less than the median log2 mean. Factors
-for each outlier gene are calculated by dividing mean expression by the
-median mean expression. A log-normal distribution is then fitted to these
-factors in order to estimate the outlier factor location and scale
-parameters. See \code{\link[fitdistrplus]{fitdist}} for details on the
-fitting.
+probability. Factors for each outlier gene are calculated by dividing mean
+expression by the median mean expression. A log-normal distribution is then
+fitted to these factors in order to estimate the outlier factor location and
+scale parameters using \code{\link[fitdistrplus]{fitdist}}.
}
diff --git a/man/winsorize.Rd b/man/winsorize.Rd
new file mode 100644
index 0000000000000000000000000000000000000000..ddef66fdb41df7ccc804cf0b4ab160de25434104
--- /dev/null
+++ b/man/winsorize.Rd
@@ -0,0 +1,19 @@
+% Generated by roxygen2: do not edit by hand
+% Please edit documentation in R/utils.R
+\name{winsorize}
+\alias{winsorize}
+\title{Winsorize vector}
+\usage{
+winsorize(x, q)
+}
+\arguments{
+\item{x}{Numeric vector to winsorize}
+
+\item{q}{Percentile to set from each end}
+}
+\value{
+Winsorized numeric vector
+}
+\description{
+Set outliers in a numeric vector to a specified percentile.
+}
diff --git a/vignettes/splatter.Rmd b/vignettes/splatter.Rmd
index 8c1a8152b67529fda8f91979425c21a05df4cb57..03eb98f35c1ec9b7494f15f0be590b1a49f53977 100644
--- a/vignettes/splatter.Rmd
+++ b/vignettes/splatter.Rmd
@@ -102,8 +102,6 @@ The parameters required for the Splat simulation are briefly described here:
distribution.
* **Expression outlier parameters**
* `out.prob` - Probability that a gene is an expression outlier.
- * `out.loProb` - Probability that an expression outlier gene is lowly
- expressed (other outliers are highly expressed).
* `out.facLoc` - Location (meanlog) parameter for the expression outlier
factor log-normal distribution.
* `out.facScale` - Scale (sdlog) parameter for the expression outlier factor