Merge pull request #245 from zhewa/SCE

Sce

.travis.yml

@@ -24,4 +24,4 @@ r_github_packages:
 after_success:
   - R CMD INSTALL $PKG_TARBALL
   - Rscript -e 'covr::coveralls()'
-  - Rscript -e 'library(lintr); lint_package(linters = with_defaults(object_length_linter(length = 40L), object_name_linter = NULL, commented_code_linter = NULL, object_usage_linter = NULL), exclusions = list("R/RcppExports.R"))'
+  - Rscript -e 'library(lintr); lint_package(linters = with_defaults(object_length_linter(length = 40L), object_name_linter = NULL, commented_code_linter = NULL, object_usage_linter = NULL, cyclocomp_linter(complexity_limit = 40)), exclusions = list("R/RcppExports.R"))'

DESCRIPTION

@@ -1,12 +1,14 @@
 Package: celda
 Title: CEllular Latent Dirichlet Allocation
-Version: 1.1.2
-Authors@R: c(person("Joshua", "Campbell", email = "camp@bu.edu", role = c("aut", "cre")),
-             person("Sean", "Corbett", email = "scorbett@bu.edu", role = c("aut")),
-	     person("Yusuke", "Koga", email="ykoga07@bu.edu", role = c("aut")),
-	     person("Zhe", "Wang", email="zhe@bu.edu", role = c("aut")))
-Description: celda leverages Bayesian hierarchical modeling to cluster genes,
-    cells, or both simultaneously from single cell sequencing data.
+Version: 1.3.1
+Authors@R: c(person("Joshua", "Campbell", email = "camp@bu.edu",
+    role = c("aut", "cre")),
+    person("Sean", "Corbett", email = "scorbett@bu.edu", role = c("aut")),
+    person("Yusuke", "Koga", email="ykoga07@bu.edu", role = c("aut")),
+    person("Shiyi", "Yang", email="syyang@bu.edu", role = c("aut")),
+    person("Eric", "Reed", email="reeder@bu.edu", role = c("aut")),
+    person("Zhe", "Wang", email="zhe@bu.edu", role = c("aut")))
+Description: celda is a Bayesian hierarchical model that can co-cluster features and cells in single cell sequencing data.
 Depends:
     R (>= 3.6)
 VignetteBuilder: knitr
@@ -31,8 +33,9 @@ Imports:
     S4Vectors,
     data.table,
     Rcpp,
+    RcppArmadillo,
     RcppEigen,
-    umap,
+    uwot,
     enrichR,
     stringi,
     SummarizedExperiment,
@@ -42,7 +45,12 @@ Imports:
     withr,
     dendextend,
     ggdendro,
-    pROC
+    pROC,
+    magrittr,
+    scater (>= 1.14.4),
+    scran,
+    SingleCellExperiment,
+    dbscan
 Suggests:
     testthat,
     knitr,
@@ -55,10 +63,10 @@ Suggests:
     M3DExampleData,
     BiocManager,
     BiocStyle
-LinkingTo: Rcpp, RcppEigen
+LinkingTo: Rcpp, RcppArmadillo, RcppEigen
 License: MIT + file LICENSE
 Encoding: UTF-8
 LazyData: false
-RoxygenNote: 6.1.1
+RoxygenNote: 7.0.2
 BugReports: https://github.com/campbio/celda/issues
 biocViews: SingleCell, GeneExpression, Clustering, Sequencing, Bayesian

NAMESPACE

@@ -1,9 +1,9 @@
 # Generated by roxygen2: do not edit by hand
 
+export("decontXcounts<-")
 export(appendCeldaList)
 export(availableModels)
 export(bestLogLikelihood)
-export(buildTreeHybrid)
 export(celda)
 export(celdaGridSearch)
 export(celdaHeatmap)
@@ -19,11 +19,13 @@ export(clusters)
 export(compareCountMatrix)
 export(countChecksum)
 export(decontX)
+export(decontXcounts)
 export(differentialExpression)
 export(distinctColors)
 export(factorizeMatrix)
 export(featureModuleLookup)
 export(featureModuleTable)
+export(findMarkers)
 export(geneSetEnrich)
 export(getDecisions)
 export(logLikelihood)
@@ -63,6 +65,7 @@ export(simulateContaminatedMatrix)
 export(subsetCeldaList)
 export(topRank)
 export(violinPlot)
+exportMethods("decontXcounts<-")
 exportMethods(bestLogLikelihood)
 exportMethods(celdaHeatmap)
 exportMethods(celdaPerplexity)
@@ -72,6 +75,8 @@ exportMethods(celdaUmap)
 exportMethods(clusterProbability)
 exportMethods(clusters)
 exportMethods(countChecksum)
+exportMethods(decontX)
+exportMethods(decontXcounts)
 exportMethods(factorizeMatrix)
 exportMethods(featureModuleLookup)
 exportMethods(logLikelihoodHistory)
@@ -91,7 +96,9 @@ import(grDevices)
 import(graphics)
 import(grid)
 import(gridExtra, except = c(combine))
+import(magrittr)
 import(stats, except = c(start, end))
+import(uwot)
 importFrom(MAST,FromMatrix)
 importFrom(MAST,summary)
 importFrom(MAST,zlm)
@@ -126,6 +133,7 @@ importFrom(gtable,gtable_height)
 importFrom(gtable,gtable_width)
 importFrom(matrixStats,logSumExp)
 importFrom(methods,.hasSlot)
+importFrom(methods,hasArg)
 importFrom(methods,is)
 importFrom(methods,new)
 importFrom(pROC,auc)
@@ -138,8 +146,6 @@ importFrom(scales,brewer_pal)
 importFrom(scales,dscale)
 importFrom(scales,hue_pal)
 importFrom(stringi,stri_list2matrix)
-importFrom(umap,umap)
-importFrom(umap,umap.defaults)
 importFrom(withr,with_seed)
 useDynLib(celda,"_colSumByGroup")
 useDynLib(celda,"_colSumByGroupChange")

R/RcppExports.R

@@ -1,10 +1,34 @@
 # Generated by using Rcpp::compileAttributes() -> do not edit by hand
 # Generator token: 10BE3573-1514-4C36-9D1C-5A225CD40393
 
+decontXEM <- function(counts, counts_colsums, theta, eta, phi, z, pseudocount) {
+    .Call('_celda_decontXEM', PACKAGE = 'celda', counts, counts_colsums, theta, eta, phi, z, pseudocount)
+}
+
+decontXLogLik <- function(counts, theta, eta, phi, z, pseudocount) {
+    .Call('_celda_decontXLogLik', PACKAGE = 'celda', counts, theta, eta, phi, z, pseudocount)
+}
+
+decontXInitialize <- function(counts, theta, z, pseudocount) {
+    .Call('_celda_decontXInitialize', PACKAGE = 'celda', counts, theta, z, pseudocount)
+}
+
+calculateNativeMatrix <- function(counts, native_counts, theta, eta, phi, z, row_index, col_index, pseudocount) {
+    .Call('_celda_calculateNativeMatrix', PACKAGE = 'celda', counts, native_counts, theta, eta, phi, z, row_index, col_index, pseudocount)
+}
+
 cG_calcGibbsProbY_Simple <- function(counts, nGbyTS, nTSbyC, nbyTS, nbyG, y, L, index, gamma, beta, delta) {
     .Call('_celda_cG_calcGibbsProbY_Simple', PACKAGE = 'celda', counts, nGbyTS, nTSbyC, nbyTS, nbyG, y, L, index, gamma, beta, delta)
 }
 
+cG_CalcGibbsProbY_ori <- function(index, counts, nTSbyC, nbyTS, nGbyTS, nbyG, y, L, nG, lg_beta, lg_gamma, lg_delta, delta) {
+    .Call('_celda_cG_CalcGibbsProbY_ori', PACKAGE = 'celda', index, counts, nTSbyC, nbyTS, nGbyTS, nbyG, y, L, nG, lg_beta, lg_gamma, lg_delta, delta)
+}
+
+cG_CalcGibbsProbY_fastRow <- function(index, counts, nTSbyC, nbyTS, nGbyTS, nbyG, y, L, nG, lg_beta, lg_gamma, lg_delta, delta) {
+    .Call('_celda_cG_CalcGibbsProbY_fastRow', PACKAGE = 'celda', index, counts, nTSbyC, nbyTS, nGbyTS, nbyG, y, L, nG, lg_beta, lg_gamma, lg_delta, delta)
+}
+
 cG_CalcGibbsProbY <- function(index, counts, nTSbyC, nbyTS, nGbyTS, nbyG, y, L, nG, lg_beta, lg_gamma, lg_delta, delta) {
     .Call('_celda_cG_CalcGibbsProbY', PACKAGE = 'celda', index, counts, nTSbyC, nbyTS, nGbyTS, nbyG, y, L, nG, lg_beta, lg_gamma, lg_delta, delta)
 }

R/all_generics.R

@@ -596,33 +596,26 @@ setGeneric("celdaTsne",
 #'  requires more memory. Default 25000.
 #' @param minClusterSize Integer. Do not subsample cell clusters below this
 #'  threshold. Default 100.
-#' @param initialDims Integer. PCA will be used to reduce the dimentionality
-#'  of the dataset. The top 'initialDims' principal components will be used
-#'  for umap. Default 20.
 #' @param modules Integer vector. Determines which features modules to use for
 #'  tSNE. If NULL, all modules will be used. Default NULL.
 #' @param seed Integer. Passed to \link[withr]{with_seed}. For reproducibility,
 #'  a default value of 12345 is used. If NULL, no calls to
 #'  \link[withr]{with_seed} are made.
-#' @param umapConfig An object of class "umapConfig" specifying parameters to
 #'  the UMAP algorithm.
-#' @return Numeric Matrix of dimension `ncol(counts)` x 2, colums representing
-#'  the "X" and "Y" coordinates in the data's t-SNE represetation.
-#' @examples
+#' @param ... Additional parameters to `uwot::umap`
+#' @return A two column matrix of UMAP coordinates#' @examples
 #' data(celdaCGSim, celdaCGMod)
-#' tsneRes <- celdaUmap(celdaCGSim$counts, celdaCGMod)
-#' @importFrom umap umap.defaults
+#' umapRes <- celdaUmap(celdaCGSim$counts, celdaCGMod)
 #' @export
 setGeneric("celdaUmap",
     signature = "celdaMod",
     function(counts,
         celdaMod,
-        maxCells = 25000,
+        maxCells = NULL,
         minClusterSize = 100,
-        initialDims = 20,
         modules = NULL,
         seed = 12345,
-        umapConfig = umap::umap.defaults) {
+        ...) {
         standardGeneric("celdaUmap")
     })
 

R/buildTreeHelper.R

@@ -319,10 +319,10 @@
     splitStats <- vapply(
         colnames(features),
         function(feat, features, class, splitMetric) {
-            splitMetric(feat, class, features, rPerf = T)
+            splitMetric(feat, class, features, rPerf = TRUE)
         }, features, class, splitMetric, FUN.VALUE = double(1))
     names(splitStats) <- colnames(features)
-    splitStats <- sort(splitStats, decreasing = T)
+    splitStats <- sort(splitStats, decreasing = TRUE)
 
     return(splitStats)
 }
@@ -412,7 +412,7 @@
     featValues <- features[, feat]
 
     # Get order of values
-    ord <- order(featValues, decreasing = T)
+    ord <- order(featValues, decreasing = TRUE)
 
     # Get sorted class and values
     featValuesSort <- featValues[ord]
@@ -516,7 +516,7 @@
     featValues <- features[, feat]
 
     # Get order of values
-    ord <- order(featValues, decreasing = T)
+    ord <- order(featValues, decreasing = TRUE)
 
     # Get sorted class and values
     featValuesSort <- featValues[ord]
@@ -835,7 +835,7 @@
                     stat = HM,
                     stringsAsFactors = F))
             }, .splitMetricModF1, fSub, cSub, group2only))
-        altStats <- altStats[order(altStats$stat, decreasing = T), ]
+        altStats <- altStats[order(altStats$stat, decreasing = TRUE), ]
 
         # Get alternative splits
         splitStats <- altStats$stat[1]