Merge pull request #9 from atuldeshpande/main

Merging software updates from atuldeshpande/SpaceMarkers

.Rbuildignore

@@ -0,0 +1,5 @@
+Dockerfile
+.gitignore
+.github
+.dockerignore
+/vignettes/test-args.Rmd

.dockerignore

@@ -0,0 +1,3 @@
+pat.txt
+.DS_Store
+.github

.github/workflows/main.yml

@@ -0,0 +1,11 @@
+# Shows how the r-build-check.yml can be called from another repository.
+name: build-check
+on:
+    pull_request:
+        branches:
+        - 'devel'
+        - 'main'
+    workflow_dispatch:
+jobs:
+    build-check:
+        uses: FertigLab/actions/.github/workflows/[email protected]

.gitignore

@@ -19,6 +19,7 @@
 
 # RStudio files
 .Rproj.user/
+SpaceMarkers.Rproj
 
 # produced vignettes
 vignettes/*.html
@@ -37,3 +38,16 @@ vignettes/*.pdf
 
 # R Environment Variables
 .Renviron
+.DS_Store
+tests/testthat/.DS_Store
+tests/testthat/testdata/.DS_Store
+vignettes/.DS_Store
+
+# GitHub directories
+.github/
+
+# Images
+*.png
+
+# GH token
+pat.txt

DESCRIPTION

@@ -1,15 +1,58 @@
-Package: SpaceMarkers
 Type: Package
+Package: SpaceMarkers
 Title: Spatial Interaction Markers
-Version: 0.1.0
-Author: Atul Deshpande [aut, cre], 
-	Ludmila Danilova [ctb]
-Maintainer: Ludmila Danilova <[email protected]>
-Description: An R/Bioconductor software tool to identify genes associated with latent space interactions in spatial transcriptomics.
-Depends: ape, dplyr, hdf5r, jsonlite, matrixStats, matrixTests, pracma, rstatix, spatstat
-Imports: Matrix, qvalue, spatstat.core, spatstat.geom, stats, utils
-License: MIT + file LICENSE
+Version: 0.99.8
+Authors@R: c(
+    person(given = "Atul", family = "Deshpande", email = "[email protected]", role = c("aut", "cre"), comment = c(ORCID="0000-0001-5144-6924")),
+    person(given = "Ludmila", family = "Danilova", email = "[email protected]", role = "ctb"),
+    person(given = "Dimitrijs", family = "Lvovs", email = "[email protected]", role = "ctb")
+  )
+BugReports: https://github.com/atuldeshpande/SpaceMarkers/issues
+URL: https://github.com/atuldeshpande/SpaceMarkers
+Description: Spatial transcriptomic technologies have helped to resolve the connection
+            between gene expression and the 2D orientation of tissues relative to each other.
+            However, the limited single-cell resolution makes it difficult to highlight the most 
+            important molecular interactions in these tissues. SpaceMarkers, R/Bioconductor software,
+            can help to find molecular interactions, by identifying genes associated with latent space 
+            interactions in spatial transcriptomics.
+Depends: 
+    R (>= 4.4.0)
+biocViews: SingleCell, GeneExpression, Software, Spatial, Transcriptomics
+Imports:
+    matrixStats,
+    matrixTests,
+    rstatix,
+    spatstat.explore,
+    spatstat.geom,
+    ape,
+    hdf5r,
+    jsonlite,
+    Matrix,
+    qvalue,
+    stats,
+    utils,
+    methods
+Suggests: 
+    data.table,
+    devtools,
+    dplyr,
+    ggplot2,
+    hrbrthemes,
+    knitr,
+    RColorBrewer,
+    cowplot,
+    readbitmap,
+    rjson,
+    rmarkdown,
+    BiocStyle,
+    testthat (>= 3.0.0),
+    viridis,
+    CoGAPS
+VignetteBuilder: 
+    knitr
+Config/testthat/edition: 3
 Encoding: UTF-8
-LazyData: true
+LazyData: false
 Roxygen: list(markdown = TRUE)
-RoxygenNote: 7.1.2
+RoxygenNote: 7.3.1
+License: MIT + file LICENSE

Dockerfile

@@ -0,0 +1,25 @@
+#generated from 69d000eee2de41886a5732a1436be4a98f080dcf
+# --platform=linux/amd64 to avoid 'no match for platform in the manifest' on M1
+FROM rocker/tidyverse:4
+
+COPY . /spacemarkers
+WORKDIR /spacemarkers
+
+RUN sudo apt-get update -y && \
+    apt-get upgrade -y && \
+    apt-get install libhdf5-dev build-essential patch -y
+
+RUN Rscript -e 'devtools::install_deps()'
+
+#https://github.com/r-lib/devtools/issues/2395
+RUN Rscript -e 'devtools::install()'
+
+#optoional packages - for vizualization
+RUN Rscript -e 'install.packages("patchwork");\
+                install.packages("ggplot2");\
+                install.packages("dplyr");\
+                install.packages("DT");\
+                install.packages("Matrix");\
+                install.packages("jsonlite");\
+                install.packages("pheatmap");\
+                install.packages("RColorBrewer")'

LICENSE

@@ -1,21 +1,2 @@
-MIT License
-
-Copyright (c) 2022 atuldeshpande
-
-Permission is hereby granted, free of charge, to any person obtaining a copy
-of this software and associated documentation files (the "Software"), to deal
-in the Software without restriction, including without limitation the rights
-to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
-copies of the Software, and to permit persons to whom the Software is
-furnished to do so, subject to the following conditions:
-
-The above copyright notice and this permission notice shall be included in all
-copies or substantial portions of the Software.
-
-THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
-IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
-FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
-AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
-LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
-OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
-SOFTWARE.
+YEAR: 2024
+COPYRIGHT HOLDER: SpaceMarkers authors

LICENSE.md

@@ -0,0 +1,21 @@
+# MIT License
+
+Copyright (c) 2024 SpaceMarkers authors
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

NAMESPACE

@@ -1,17 +1,25 @@
 # Generated by roxygen2: do not edit by hand
 
 export(getInteractingGenes)
+export(getSpatialFeatures)
 export(getSpatialParameters)
 export(load10XCoords)
 export(load10XExpr)
-import(ape)
-import(dplyr)
-import(hdf5r)
-import(jsonlite)
-import(matrixStats)
-import(matrixTests)
-import(pracma)
-import(rstatix)
-import(spatstat)
-import(stats)
-import(utils)
+importFrom(Matrix,sparseMatrix)
+importFrom(ape,Moran.I)
+importFrom(hdf5r,h5file)
+importFrom(jsonlite,read_json)
+importFrom(matrixStats,count)
+importFrom(matrixTests,row_kruskalwallis)
+importFrom(methods,slot)
+importFrom(qvalue,qvalue)
+importFrom(rstatix,filter)
+importFrom(spatstat.explore,Smooth)
+importFrom(spatstat.geom,marks)
+importFrom(spatstat.geom,owin)
+importFrom(spatstat.geom,ppp)
+importFrom(stats,dist)
+importFrom(stats,lag)
+importFrom(stats,pnorm)
+importFrom(stats,sd)
+importFrom(utils,read.csv)

NEWS.md

@@ -0,0 +1,3 @@
+# SpaceMarkers 0.1.0
+
+* Added a `NEWS.md` file to track changes to the package.

R/data.R

@@ -0,0 +1,29 @@
+#' Optimal paramters of 5 patterns from CoGAPS.
+#'
+#' A dataset with the optimal width of the gaussian distribution (sigmaOpt) and
+#' the outlier threshold around the set of spots (thresOpt) for each pattern
+#' obtained from CoGAPS. CoGAPS was ran on spatial transcriptomic data from a 
+#' breast cancer sample.
+#' 
+#' @name optParams
+#' @format A data frame with 2 rows and 5 columns:
+#' \describe{
+#'   \item{Pattern_1}{immune cell pattern paramters }
+#'   \item{Pattern_2}{Disp.1 parameters}
+#'   \item{Pattern_3}{intraductal carcinoma (DCIS) parameters }
+#'   \item{Pattern_4}{Disp.2 parameters }
+#'   \item{Pattern_5}{invasive carcinoma lesion pattern paramters }
+#' }
+#' @return A matrix of optimal parameters for patterns identified by CoGAPS
+NULL
+
+#' Curated Genes for example purposes
+#' 
+#' A vector with genes selected based on previous runs of SpaceMarkers on the
+#' Visium 10x breast ductal carcinoma spatial transcriptomics dataset
+#'
+#' @name curated_genes
+#' @format A vector with 114 pre-selected genes
+#' @return a vector of genes
+NULL
+

R/find_genes_of_interest.R

@@ -0,0 +1,129 @@
+#' @importFrom matrixTests row_kruskalwallis
+#' @importFrom rstatix filter
+#' @importFrom matrixStats count
+#' @importFrom stats lag dist pnorm sd
+#' @importFrom qvalue qvalue
+## author: Atul Deshpande
+## email: [email protected]
+
+row.dunn.test <- function(in.data,region){
+    region <- factor(region)
+    pattern1 <- levels(region)[2]
+    pattern2 <- levels(region)[3]
+    in.ranks <- matrixStats::rowRanks(in.data,cols = !is.na(region),
+                                        ties.method = "average")
+    rsub <- region[!is.na(region)]
+
+    N <- length(rsub)
+    N1 <- sum(rsub==pattern1)
+    N2 <- sum(rsub==pattern2)
+    NI <- sum(rsub=="Interacting")
+
+    tiesStat <- apply(in.ranks,1,function(rr) sum(table(rr)^3-table(rr)))
+    tiesStat2 <- sqrt(1 - tiesStat/N/(N-1)/(N+1))
+
+    SEI1 <- sqrt(N*(N+1)/12*(1/NI+1/N1))
+    SEI2 <- sqrt(N*(N+1)/12*(1/NI+1/N2))
+    SE12 <- sqrt(N*(N+1)/12*(1/N1+1/N2))
+
+    mInt<- apply(in.ranks[,rsub=="Interacting"],1,mean)
+    mP1<- apply(in.ranks[,rsub==pattern1],1,mean)
+    mP2<- apply(in.ranks[,rsub==pattern2],1,mean)
+
+    zP1_Int <- (mP1-mInt)/SEI1/tiesStat2
+    zP2_Int <- (mP2-mInt)/SEI2/tiesStat2
+    zP2_P1 <- (mP2-mP1)/SE12/tiesStat2
+    zVals <-cbind(zP1_Int,zP2_Int,zP2_P1)
+
+    pval_1_Int <- ifelse(2*(pnorm(zP1_Int,lower.tail = TRUE))
+                        <=1,2*(pnorm(zP1_Int,lower.tail = TRUE)),1)
+    pval_2_Int <- ifelse(2*(pnorm(zP2_Int,lower.tail = TRUE))
+                            <=1,2*(pnorm(zP2_Int,lower.tail = TRUE)),1)
+    pval_2_1 <- ifelse(zP2_P1>0,2*(pnorm(zP2_P1,lower.tail = FALSE)),
+                        2*(pnorm(zP2_P1,lower.tail = TRUE)))
+    pvals<-cbind(pval_1_Int,pval_2_Int,pval_2_1)
+return(cbind(zVals,pvals))
+}
+
+#===================
+#' find_genes_of_interest
+#' Identify genes associated with pattern interaction.
+
+#' This function identifies genes exhibiting significantly higher values of 
+#' testMat in the Interaction region of the two 
+#' patterns compared #' to regions with exclusive influence from either 
+#' pattern. It uses Kruskal-Wallis test followed by
+#' posthoc analysis using Dunn's Test to identify the genes.
+#'
+#' @usage
+#' find_genes_of_interest(
+#'     testMat,
+#'     goodGenes = NULL,
+#'     region,
+#'     fdr.level = 0.05,
+#'     analysis = c("enrichment", "overlap")
+#' )
+#' @param    testMat A matrix of counts with cells as columns and genes as rows
+#' @param    goodGenes A vector of user specified genes expected to interact 
+#' a priori. The default for this is NULL as the function can find these genes 
+#' itself
+#' @param    region A data frame of the reference pattern regions that overlap 
+#' with the other patterns
+#' @param    fdr.level False Discovery Rate. The default value is 0.05.
+#' @param    analysis a character string that specifies the type of analysis to 
+#' carry out, whether overlap or enrichment.
+#'
+#' @return a list of genes exhibiting significantly higher values of testMat in 
+#' the Interaction region of the two #' patterns compared to regions with 
+#' exclusive influence from either pattern.
+
+
+
+find_genes_of_interest<-function(
+        testMat,goodGenes=NULL,region, fdr.level=0.05,
+        analysis=c("enrichment","overlap")) {
+    region <- factor(region)
+    patnames <- levels(region)[which(levels(region)!="Interacting")]
+    pattern1 <- patnames[1]
+    pattern2 <- patnames[2]
+    if (!is.null(goodGenes)){
+        subset_goodGenes <- intersect(rownames(testMat),goodGenes)
+        testMat <- testMat[subset_goodGenes,]
+        }
+    res_kruskal<- matrixTests::row_kruskalwallis(x=as.matrix(testMat),g=region)
+    qq <- qvalue::qvalue(res_kruskal$pvalue,fdr.level = fdr.level,
+                            pfdr = FALSE, pi0 = 1)
+    res_kruskal <- cbind(res_kruskal,p.adj = qq$qvalues)
+    res_dunn_test <- row.dunn.test(as.matrix(testMat),region)
+    rownames(res_dunn_test) <- rownames(res_kruskal)
+    ind <- rownames(res_kruskal[which(res_kruskal$p.adj<fdr.level),])
+    qDunn <- qvalue::qvalue(res_dunn_test[,4:6],
+                            fdr.level = fdr.level, pfdr = FALSE, pi0 = 1)
+    qq<-qvalue::qvalue(
+        res_dunn_test[ind,4:6],fdr.level=fdr.level,pfdr=FALSE,pi0 = 1)
+    qDunn$qvalues[ind,] <- qq$qvalue
+    res_dunn_test <- cbind(res_dunn_test,qDunn$qvalues)
+    colnames(res_dunn_test)[7:9] <- paste0(colnames(res_dunn_test)[7:9],".adj")
+    interact_patt1 <- res_dunn_test[ind,"pval_1_Int.adj"]<fdr.level
+    interact_patt2 <- res_dunn_test[ind,"pval_2_Int.adj"]<fdr.level
+    interacting_over_both_patterns <- interact_patt1 & interact_patt2
+    not_pattern1_diff_pattern2 <- res_dunn_test[ind,"pval_2_1.adj"]>=fdr.level
+    exc_interact_patt1 <- interact_patt1 & not_pattern1_diff_pattern2
+    exc_interact_patt2 <- interact_patt2 & not_pattern1_diff_pattern2
+    names(exc_interact_patt2)<-ind
+    names(exc_interact_patt1)<-names(exc_interact_patt2)
+    names(interacting_over_both_patterns)<-names(exc_interact_patt1)
+    interact_genes<-matrix(FALSE,nrow=nrow(res_dunn_test),ncol=2,dimnames=list(
+        rownames(res_dunn_test),c("Gene",paste0(pattern1,' x ',pattern2))))
+    interact_genes[,1] <- rownames(interact_genes)
+    interact_genes[ind[which(exc_interact_patt1)],2]<-paste0("vs",pattern1)
+    interact_genes[ind[which(exc_interact_patt2)],2]<-paste0("vs",pattern2)
+    interact_genes[ind[which(interacting_over_both_patterns)],2]<-"vsBoth"
+    colnames(res_kruskal) <- paste0("KW.",colnames(res_kruskal))
+    colnames(res_dunn_test) <- paste0("Dunn.",colnames(res_dunn_test))
+    interact_genes <- cbind(interact_genes,res_kruskal,res_dunn_test)
+    if (analysis=="overlap"){
+        interact_genes <-interact_genes[interact_genes[,2]!="FALSE",] 
+    }
+    return(list(interact_genes))
+}