updated simulation data

.gitignore

@@ -1,34 +1,35 @@
-tags
-test/*
-.pdf
-
-.RData
-.Rhistory
-*~
-.*~
-.DS_Store
-# Byte-compiled / optimized / DLL files
-__pycache__/
-*.py[cod]
-*$py.class
-
-# C extensions
-*.so
-
-# Distribution / packaging
-.Python
-build/
-develop-eggs/
-dist/
-downloads/
-eggs/
-.eggs/
-lib/
-lib64/
-parts/
-sdist/
-var/
-wheels/
+tags
+test/*
+*.pdf
+.RData
+.Rhistory
+*~
+*csv
+*tsv
+.*~
+.DS_Store
+# Byte-compiled / optimized / DLL files
+__pycache__/
+*.py[cod]
+*$py.class
+
+# C extensions
+*.so
+
+# Distribution / packaging
+.Python
+build/
+develop-eggs/
+dist/
+downloads/
+eggs/
+.eggs/
+lib/
+lib64/
+parts/
+sdist/
+var/
+wheels/
 pip-wheel-metadata/
 share/python-wheels/
 *.egg-info/

metrics/data-sim/call-deconv-on-sim.cwl

@@ -1,109 +1,109 @@
-#!/usr/bin/env cwltool
-class: Workflow
-label: call-deconv-on-sim
-id: call-deconv-on-sim
-cwlVersion: v1.2
-
-requirements:
-  - class: SubworkflowFeatureRequirement
-  - class: MultipleInputFeatureRequirement
-  - class: InlineJavascriptRequirement
-  - class: StepInputExpressionRequirement
-
-inputs:
-   signature:  ##name of matrix to sample from
-     type: string
-   protAlg:    ##algorithm to run
-     type: string
-   simulation: ## permutation to test
-     type: string
-     default: '1'
-   dataType:   ##mRNA or protein data
-     type: string
-     default: 'prot'
-   simType:    ##data simulated from mrna or protein
-     type: string
-     default: 'prot'
-   sample:     ##how much of the permuted sample do we test
-     type: int
-     default: 100
-   sampleRep:  ##
-     type: int
-     default: 1
-
-outputs:
-#  matrix:
-#     type: File
-#     outputSource: get-sim-data/matrix
-#  cellPred:
-#     type: File
-#     outputSource: get-sim-data/cellType
-#  deconvoluted:
-#     type: File
-#     outputSource: deconv-prot/deconvoluted
-#  deconv:
-#     type: File
-#     outputSource: match-prot-to-sig/updated-deconv
-  cell-cor-file:
-     type: File
-     outputSource: celltype-cor/corr
-
-steps:
-  get-sig-mat:
-     run: ../../signature_matrices/get-signature-matrix.cwl
-     in:
-      sigMatrixName: signature
-      subsample: sample
-     out:
-      [sigMatrix]
-  get-sim-data:
-     run: ../../simulatedData/sim-data-tool.cwl
-     in:
-       simNumber: simulation
-       simType: simType
-     out:
-       [matrix,cellType]
-  deconv-prot:
-     run: ../../tumorDeconvAlgs/run-deconv.cwl
-     in:
-       alg: protAlg
-       signature: get-sig-mat/sigMatrix
-       matrix: get-sim-data/matrix
-     out: [deconvoluted]
-  match-prot-to-sig:
-     run: ../../simulatedData/map-sig-tool.cwl
-     in:
-       deconv-matrix: deconv-prot/deconvoluted
-       sig-matrix: get-sig-mat/sigMatrix
-       deconv-type: simType
-       cell-matrix: get-sim-data/cellType
-     out: [updated-deconv,updated-cell-matrix]
-  celltype-cor:
-     run: ../correlations/deconv-corrXcelltypes-cwl-tool.cwl
-     in:
-       cancerType: simulation
-       mrnaAlg:
-          valueFrom: "cellFraction"
-       protAlg: protAlg
-       signature: signature
-       sampleVal: sample
-       sampleType: simType
-       sampleRep: sampleRep
-       proteomics:
-         source: match-prot-to-sig/updated-deconv
-       transcriptomics:
-         source: match-prot-to-sig/updated-cell-matrix
-     out: [corr]
-#  matrix-distance:
-#     run: ../distance/deconv-comparison-tool.cwl
-#     in:
-#       matrixA: match-prot-to-sig/updated-deconv
-#       matrixB: match-prot-to-sig/updated-cell-matrix 
-#       cancerType: simulation
-#       aAlg: protAlg
-#       bAlg:
-#         valueFrom: "cellFraction"
-#       signature: get-sig-mat/sigMatrix
-#       sampleType: simType
-#     out:
-#       [dist]
+#!/usr/bin/env cwltool
+class: Workflow
+label: call-deconv-on-sim
+id: call-deconv-on-sim
+cwlVersion: v1.2
+
+requirements:
+  - class: SubworkflowFeatureRequirement
+  - class: MultipleInputFeatureRequirement
+  - class: InlineJavascriptRequirement
+  - class: StepInputExpressionRequirement
+
+inputs:
+   signature:  ##name of matrix to sample from
+     type: string
+   protAlg:    ##algorithm to run
+     type: string
+   simulation: ## permutation to test
+     type: string
+     default: '1'
+   dataType:   ##mRNA or protein data
+     type: string
+     default: 'prot'
+   simType:    ##data simulated from mrna or protein
+     type: string
+     default: 'prot'
+   sample:     ##how much of the permuted sample do we test
+     type: int
+     default: 100
+   sampleRep:  ##
+     type: int
+     default: 1
+
+outputs:
+#  matrix:
+#     type: File
+#     outputSource: get-sim-data/matrix
+  cellPred:
+     type: File
+     outputSource: match-prot-to-sig/updated-cell-matrix
+#  deconvoluted:
+#     type: File
+#     outputSource: deconv-prot/deconvoluted
+  deconv:
+     type: File
+     outputSource: match-prot-to-sig/updated-deconv
+  cell-cor-file:
+     type: File
+     outputSource: celltype-cor/corr
+
+steps:
+  get-sig-mat:
+     run: ../../signature_matrices/get-signature-matrix.cwl
+     in:
+      sigMatrixName: signature
+      subsample: sample
+     out:
+      [sigMatrix]
+  get-sim-data:
+     run: ../../simulatedData/sim-data-tool.cwl
+     in:
+       simNumber: simulation
+       simType: simType
+     out:
+       [matrix,cellType]
+  deconv-prot:
+     run: ../../tumorDeconvAlgs/run-deconv.cwl
+     in:
+       alg: protAlg
+       signature: get-sig-mat/sigMatrix
+       matrix: get-sim-data/matrix
+     out: [deconvoluted]
+  match-prot-to-sig:
+     run: ../../simulatedData/map-sig-tool.cwl
+     in:
+       deconv-matrix: deconv-prot/deconvoluted
+       sig-matrix: get-sig-mat/sigMatrix
+       deconv-type: simType
+       cell-matrix: get-sim-data/cellType
+     out: [updated-deconv,updated-cell-matrix]
+  celltype-cor:
+     run: ../correlations/deconv-corrXcelltypes-cwl-tool.cwl
+     in:
+       cancerType: simulation
+       mrnaAlg:
+          valueFrom: "cellFraction"
+       protAlg: protAlg
+       signature: signature
+       sampleVal: sample
+       sampleType: simType
+       sampleRep: sampleRep
+       proteomics:
+         source: match-prot-to-sig/updated-deconv
+       transcriptomics:
+         source: match-prot-to-sig/updated-cell-matrix
+     out: [corr]
+#  matrix-distance:
+#     run: ../distance/deconv-comparison-tool.cwl
+#     in:
+#       matrixA: match-prot-to-sig/updated-deconv
+#       matrixB: match-prot-to-sig/updated-cell-matrix 
+#       cancerType: simulation
+#       aAlg: protAlg
+#       bAlg:
+#         valueFrom: "cellFraction"
+#       signature: get-sig-mat/sigMatrix
+#       sampleType: simType
+#     out:
+#       [dist]

metrics/data-sim/plotSampling.R

@@ -0,0 +1,16 @@
+##plot simulation results with sampling
+
+
+allfiles=list.files('.')
+corfiles=allfiles[grep('correlation-',allfiles)]
+
+fulltab<-do.call(rbind,lapply(corfiles,function(x) read.table(x,header=T)))
+
+library(ggplot2)
+
+res = ggplot(fulltab,aes(x=sample,y=value,col=cellType))+geom_jitter()+facet_grid(matrix~prot.algorithm)
+ggsave('samplingResultsForProt.pdf',res,width=12)
+
+
+##now do the mrna
+res2<-ggplot(fulltab,aes(x=sample,y=value,fill=cellType))+geom_point()+facet_grid(matrix~prot.algorithm)

metrics/data-sim/runSamplingManually.py

@@ -7,19 +7,19 @@
 
 filelist = []
 for i in [0,1,2,3,4]:
-    estring = 'cwltool simul-data-sampling.cwl --prot-sigs LM9 --simType prot --repNumber '+str(i)
-    print(estring)
-    os.system(estring)
-    fname='combined-cellType-correlation-'+str(i)+'.tsv'
-    os.rename(fname,'prot-LM9-'+fname)
-    filelist.append('prot-LM9-'+fname)
-
-    estring = 'cwltool simul-data-sampling.cwl --prot-sigs LM7c --simType prot --repNumber '+str(i)
-    print(estring)
-    os.system(estring)
-    fname='combined-cellType-correlation-'+str(i)+'.tsv'
-    os.rename(fname,'prot-LM7c-'+fname)
-    filelist.append('prot-LM7c-'+fname)
+#    estring = 'cwltool simul-data-sampling.cwl --prot-sigs LM9 --simType prot --repNumber '+str(i)
+#    print(estring)
+#    os.system(estring)
+#    fname='combined-cellType-correlation-'+str(i)+'.tsv'
+#    os.rename(fname,'prot-LM9-'+fname)
+#    filelist.append('prot-LM9-'+fname)
+
+#    estring = 'cwltool simul-data-sampling.cwl --prot-sigs LM7c --simType prot --repNumber '+str(i)
+#    print(estring)
+#    os.system(estring)
+#    fname='combined-cellType-correlation-'+str(i)+'.tsv'
+#    os.rename(fname,'prot-LM7c-'+fname)
+#    filelist.append('prot-LM7c-'+fname)
 
     estring = 'cwltool simul-data-sampling.cwl --rna-sigs LM22 --simType mrna --repNumber '+str(i)
     print(estring)

metrics/data-sim/simul-data-sampling.cwl

@@ -20,7 +20,7 @@ inputs:
       default: [10,20,40,60,80,100]
    mrna-perms:
       type: string[]
-      default: ['1','2','3','4','5']#,'6','7','8','9','10','pbmc']
+      default: ['1','2','3','4','5','6','7','8','9','10','pbmc']
    prot-perms:
       type: string[]
       default: ['1','2','3','4','5']

simulatedData/mapSimDataMatrices.R

@@ -8,12 +8,13 @@ main <- function(){
     print(argv)
     deconv.mat <- read.table(argv[1],header=T,row.names=1,sep='\t', check.names = F) #matrix to be fixed
     sig.mat <- argv[2] #signature used to deconvolve matrix
+    sig.mat<-basename(sig.mat)
     sig.mat <-stringr::str_replace(sig.mat,'_[0-9]*','') ##added this due to sampling changes!!!
     sim.type <- tolower(argv[3]) #type of simulation performed
     cell.type <-read.table(argv[4],header=T,row.names=1,sep='\t', check.names = F)
 
-    sig.mat<-basename(sig.mat)
 
+    print(paste("mapping",sig.mat,"predictions to",sim.type,'simulations'))
     over <- intersect(rownames(deconv.mat),rownames(cell.type))
     print(paste0('Sig cell types: ',paste(rownames(deconv.mat),collapse=',')))
     print(paste0('Simulated cell types: ',paste(rownames(cell.type),collapse=',')))
@@ -24,7 +25,7 @@ main <- function(){
     ##maximize overlap
     rownames(deconv.mat)<-rownames(deconv.mat)%>%
         stringr::str_replace('Neutrophil$','Neutrophils')%>%
-        stringr::str_replace(fixed('MO'),'Monocytes')%>%
+        stringr::str_replace(stringr::fixed('MO'),'Monocytes')%>%
         stringr::str_replace('^T8 cells$','CD8 T cells')%>%
         stringr::str_replace('^T4 cells$','CD4 T cells')%>%
         stringr::str_replace(fixed('B-cells'),'B cells')%>%