Skip to content

Commit

Permalink
fix scripts parameters
Browse files Browse the repository at this point in the history
  • Loading branch information
Sergii Domanskyi committed Mar 12, 2022
1 parent a176be9 commit e45e1d6
Show file tree
Hide file tree
Showing 5 changed files with 60 additions and 53 deletions.
63 changes: 35 additions & 28 deletions bin/characterization_BayesSpace.R
View file
Original file line number Diff line number Diff line change
Expand Up @@ -6,36 +6,36 @@ library(SingleCellExperiment)
library(ggplot2)
library(BayesSpace)
library(Matrix)
library(reticulate)


# Parse command-line arguments
parser <- ArgumentParser()

args <- parser$add_argument_group("Agruments", "required and optional arguments")

args$add_argument("--filePath", help="Path to npz counts file", metavar="dir", required=TRUE)
args$add_argument("--nameX", default="st_adata_X.npz", help="Path to X", metavar="file", required=FALSE)
args$add_argument("--nameVar", default="st_adata.var.csv", help="Path to features metadata", metavar="file", required=FALSE)
args$add_argument("--nameObs", default="st_adata.obs.csv", help="Path to observation metadata", metavar="file", required=FALSE)
args$add_argument("--countsFactor", default=100, help="factor", metavar="factor", required=FALSE)
args$add_argument("--numberHVG", default=2000, help="factor", metavar="factor", required=FALSE)
args$add_argument("--numberPCs", default=7, help="factor", metavar="factor", required=FALSE)
args$add_argument("--minClusters", default=2, help="factor", metavar="factor", required=FALSE)
args$add_argument("--maxClusters", default=10, help="factor", metavar="factor", required=FALSE)
args$add_argument("--optimalQ", default=5, help="factor", metavar="factor", required=FALSE)
args$add_argument("--STplatform", default="Visium", help="Technology grid", metavar="factor", required=FALSE)

library(reticulate)


# Parse command-line arguments
parser <- ArgumentParser()

args <- parser$add_argument_group("Agruments", "required and optional arguments")

args$add_argument("--filePath", help="Path to npz counts file", metavar="dir", required=TRUE)
args$add_argument("--nameX", default="st_adata_X.npz", help="Path to X", metavar="file", required=FALSE)
args$add_argument("--nameVar", default="st_adata.var.csv", help="Path to features metadata", metavar="file", required=FALSE)
args$add_argument("--nameObs", default="st_adata.obs.csv", help="Path to observation metadata", metavar="file", required=FALSE)
args$add_argument("--countsFactor", default=100, help="factor", metavar="factor", required=FALSE)
args$add_argument("--numberHVG", default=2000, type="integer", help="factor", metavar="factor", required=FALSE)
args$add_argument("--numberPCs", default=7, type="integer", help="factor", metavar="factor", required=FALSE)
args$add_argument("--minClusters", default=2, type="integer", help="factor", metavar="factor", required=FALSE)
args$add_argument("--maxClusters", default=10, type="integer", help="factor", metavar="factor", required=FALSE)
args$add_argument("--optimalQ", default=5, type="integer", help="factor", metavar="factor", required=FALSE)
args$add_argument("--STplatform", default="Visium", help="Technology grid", metavar="factor", required=FALSE)

args$add_argument("--qtuneSaveName", default="st_bayes_qtune.png", help="file name", metavar="file", required=FALSE)
args$add_argument("--bayesClustersName", default="st_bayes_clusters.png", help="file name", metavar="file", required=FALSE)
args$add_argument("--bayesClustersEnhancedName", default="st_bayes_clusters_enhanced.png", help="file name", metavar="file", required=FALSE)
args$add_argument("--bayesOriginalEnhancedFeatures", default="st_bayes_original_and_enhanced.png", help="file name", metavar="file", required=FALSE)
args$add_argument("--bayesEnhancedMarkers", default="bayes_enhanced_markers.csv", help="file name", metavar="file", required=FALSE)
args$add_argument("--bayesSubspotCoord", default="bayes_subspot_cluster_and_coord.csv", help="file name", metavar="file", required=FALSE)
args$add_argument("--bayesSpotCluster", default="bayes_spot_cluster.csv", help="file name", metavar="file", required=FALSE)

args <- parser$parse_args()

args$add_argument("--bayesSpotCluster", default="bayes_spot_cluster.csv", help="file name", metavar="file", required=FALSE)

args <- parser$parse_args()


# #### Spatial clustering:
# The latent cluster is modeled to depend on three parameters: cluster mean, fixed precision matrix and scaling factor. <br>
Expand All @@ -52,8 +52,8 @@ args <- parser$parse_args()

# #### Tuning number of clusters:
# Elbow of negative pseudo-log-likelihood curve, that estimates how well the model fit the data for each number of clusters. <br>




# Main script
set.seed(123)
np <- import("numpy")
Expand All @@ -80,6 +80,13 @@ count.data <- np$load(paste0(normDataDir, args$nameX))[['arr_0']] * args$countsF
colnames(count.data) <- st_obs_all$X
rownames(count.data) <- rowData

print(args$countsFactor)
print(args$STplatform)
print(args$numberPCs)
print(args$numberHVG)
print(args$minClusters)


dsp <- SingleCellExperiment(assays=list(counts=count.data), rowData=row_df, colData=col_df)
dsp <- spatialPreprocess(dsp, platform=args$STplatform, n.PCs=args$numberPCs, n.HVGs=args$numberHVG, log.normalize=TRUE)

Expand All @@ -92,7 +99,7 @@ dsp <- spatialCluster(dsp, q=args$optimalQ, platform=args$STplatform, d=args$num

clusterPlot(dsp, palette=c("purple", "cyan", "blue", "yellow", "red"), color=NA) + theme_bw() + xlab("Column") + ylab("Row") + labs(fill="BayesSpace\ncluster", title="Spatial clustering")
ggsave(paste0(normDataDir, args$bayesClustersName), dpi=600, scale=0.75, width=8, height=8, units="in")


dsp.enhanced <- spatialEnhance(dsp, q=args$optimalQ, platform="Visium", d=args$numberPCs, model="t", gamma=2, jitter_prior=0.3, jitter_scale=3.5, nrep=1000, burn.in=100)

clusterPlot(dsp.enhanced, palette=c("purple", "cyan", "blue", "yellow", "red"), color=NA) + theme_bw() + xlab("Column") + ylab("Row") + labs(fill="BayesSpace\ncluster", title="Spatial clustering")
Expand All @@ -119,4 +126,4 @@ write.csv(df_enhanced_markers, file=paste0(args$filePath, args$bayesEnhancedMark
df_spot_cluster <- subset(as.data.frame(dsp@colData), select=-c(imagecol, imagerow))
write.csv(df_spot_cluster, file=paste0(args$filePath, args$bayesSpotCluster))

quit(status=0)
quit(status=0)
12 changes: 6 additions & 6 deletions bin/characterization_SPOTlight.R
View file
Original file line number Diff line number Diff line change
Expand Up @@ -34,17 +34,17 @@ args$add_argument("--SCnameObs", default="sc_adata.obs.csv", help="Path to obser
args$add_argument("--fileh5", default="raw_feature_bc_matrix.h5", help="File HDF5", metavar="file", required=FALSE) # "filtered_feature_bc_matrix.h5"

args$add_argument("--outsSubDir", default="raw_feature_bc_matrix/", help="dir", metavar="file", required=FALSE)
args$add_argument("--mtxGeneColumn", default=2, help="columns index", metavar="col", required=FALSE)
args$add_argument("--countsFactor", default=100, help="factor", metavar="factor", required=FALSE)
args$add_argument("--mtxGeneColumn", default=2, type="integer", help="columns index", metavar="col", required=FALSE)
args$add_argument("--countsFactor", default=100, type="integer", help="factor", metavar="factor", required=FALSE)

args$add_argument("--clusterResolution", default=0.3, help="factor", metavar="factor", required=FALSE)
args$add_argument("--clusterResolution", default=0.3, type="double", help="factor", metavar="factor", required=FALSE)

args$add_argument("--numberHVG", default=3000, help="factor", metavar="factor", required=FALSE)
args$add_argument("--numberCellsPerCelltype", default=100, help="factor", metavar="factor", required=FALSE)
args$add_argument("--numberHVG", default=3000, type="integer", help="factor", metavar="factor", required=FALSE)
args$add_argument("--numberCellsPerCelltype", default=100, type="integer", help="factor", metavar="factor", required=FALSE)
args$add_argument("--NMFsaveFile", default="SPOTlight_ls_mk_normed.rds", help="File to save NMF RDS", metavar="file", required=FALSE)

args$add_argument("--SPOTlightScatterpiesName", default="SPOTlight_st_scatterpies.png", help="dir", metavar="file", required=FALSE)
args$add_argument("--SPOTlightScatterpiesSize", default=0.35, help="dir", metavar="file", required=FALSE)
args$add_argument("--SPOTlightScatterpiesSize", default=0.35, type="double", help="dir", metavar="file", required=FALSE)
args$add_argument("--SPOTlightSCclustersName", default="SPOTlight_sc_clusters.png", help="dir", metavar="file", required=FALSE)
args$add_argument("--SPOTlightTopicsName", default="SPOTlight_st_topic_profiles.png", help="dir", metavar="file", required=FALSE)
args$add_argument("--SPOTlightFeaturesName", default="SPOTlight_st_prop.png", help="dir", metavar="file", required=FALSE)
Expand Down
16 changes: 8 additions & 8 deletions bin/characterization_STdeconvolve.R
View file
Original file line number Diff line number Diff line change
Expand Up @@ -29,19 +29,19 @@ args$add_argument("--SCnameObs", default="sc_adata.obs.csv", help="Path to obser
args$add_argument("--fileh5", default="raw_feature_bc_matrix.h5", help="File HDF5", metavar="file", required=FALSE) # "filtered_feature_bc_matrix.h5"

args$add_argument("--outsSubDir", default="raw_feature_bc_matrix/", help="dir", metavar="file", required=FALSE)
args$add_argument("--mtxGeneColumn", default=2, help="columns index", metavar="col", required=FALSE)
args$add_argument("--countsFactor", default=100, help="factor", metavar="factor", required=FALSE)
args$add_argument("--mtxGeneColumn", default=2, type="integer", help="columns index", metavar="col", required=FALSE)
args$add_argument("--countsFactor", default=100, type="integer", help="factor", metavar="factor", required=FALSE)

args$add_argument("--corpusRemoveAbove", default=1.0, help="factor", metavar="factor", required=FALSE)
args$add_argument("--corpusRemoveBelow", default=0.05, help="factor", metavar="factor", required=FALSE)
args$add_argument("--corpusRemoveAbove", default=1.0, type="double", help="factor", metavar="factor", required=FALSE)
args$add_argument("--corpusRemoveBelow", default=0.05, type="double", help="factor", metavar="factor", required=FALSE)

args$add_argument("--LDAminTopics", default=8, help="factor", metavar="factor", required=FALSE)
args$add_argument("--LDAmaxTopics", default=9, help="factor", metavar="factor", required=FALSE)
args$add_argument("--LDAminTopics", default=8, type="integer", help="factor", metavar="factor", required=FALSE)
args$add_argument("--LDAmaxTopics", default=9, type="integer", help="factor", metavar="factor", required=FALSE)
args$add_argument("--LDAsaveFile", default="STdeconvolve_optLDA.rds", help="File to save LDA RDS", metavar="file", required=FALSE)

args$add_argument("--STdeconvolveScatterpiesName", default="STdeconvolve_st_scatterpies.png", help="dir", metavar="file", required=FALSE)
args$add_argument("--STdeconvolveScatterpiesSize", default=2.85, help="dir", metavar="file", required=FALSE)
args$add_argument("--STdeconvolveFeaturesSizeFactor", default=1.0, help="dir", metavar="file", required=FALSE)
args$add_argument("--STdeconvolveScatterpiesSize", default=2.85, type="double", help="dir", metavar="file", required=FALSE)
args$add_argument("--STdeconvolveFeaturesSizeFactor", default=1.0, type="double", help="dir", metavar="file", required=FALSE)
args$add_argument("--STdeconvolveFeaturesName", default="STdeconvolve_st_prop.png", help="dir", metavar="file", required=FALSE)
args$add_argument("--STdeconvolveCorrName", default="STdeconvolve_st_prop_corr.png", help="dir", metavar="file", required=FALSE)
args$add_argument("--STdeconvolvePropNormName", default="STdeconvolve_prop_norm.csv", help="dir", metavar="file", required=FALSE)
Expand Down
4 changes: 2 additions & 2 deletions bin/stSpatialDE.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -17,8 +17,8 @@
parser.add_argument('--fileName', metavar='file', type=str, default="st_adata_norm.h5ad", help='File name.')
parser.add_argument('--saveFileName', metavar='file', type=str, default="stSpatialDE.csv", help='File name.')
parser.add_argument('--savePlotName', metavar='plot', type=str, default="stSpatialDE.png", help='File name.')
parser.add_argument('--plotTopHVG', metavar='number', type=str, default=15, help='File name.')
parser.add_argument('--numberOfColumns', metavar='number', type=str, default=5, help='File name.')
parser.add_argument('--plotTopHVG', metavar='number', type=int, default=15, help='File name.')
parser.add_argument('--numberOfColumns', metavar='number', type=int, default=5, help='File name.')

args = parser.parse_args()

Expand Down
18 changes: 9 additions & 9 deletions modules/local/tasks.nf
View file
Original file line number Diff line number Diff line change
Expand Up @@ -61,9 +61,9 @@ import groovy.json.JsonSlurper
[ ! -d \${dname} ] && mkdir \${dname}
python $projectDir/bin/script_read_st_data.py --outsPath=${sample_info.st_data_dir} --saveFile=\${dname}/st_adata_raw.h5ad --countsFile=raw_feature_bc_matrix.h5 --npCountsOutputName=st_adata_counts_in_tissue.npz --minCounts=params.STload_minCounts --minCells=params.STload_minCells
python $projectDir/bin/script_read_st_data.py --outsPath=${sample_info.st_data_dir} --saveFile=\${dname}/st_adata_raw.h5ad --countsFile=raw_feature_bc_matrix.h5 --npCountsOutputName=st_adata_counts_in_tissue.npz --minCounts=$params.STload_minCounts --minCells=$params.STload_minCells
python $projectDir/bin/script_read_sc_data.py --outsPath=${sample_info.sc_data_dir} --saveFile=\${dname}/sc_adata_raw.h5ad --npCountsOutputName=sc_adata_counts.npz --minCounts=params.SCload_minCounts --minCells=params.SCload_minCells --minGenes=params.SCload_minGenes
python $projectDir/bin/script_read_sc_data.py --outsPath=${sample_info.sc_data_dir} --saveFile=\${dname}/sc_adata_raw.h5ad --npCountsOutputName=sc_adata_counts.npz --minCounts=$params.SCload_minCounts --minCells=$params.SCload_minCells --minGenes=$params.SCload_minGenes
if [[ -s \${dname}/st_adata_raw.h5ad ]] && \
[[ -s \${dname}/sc_adata_raw.h5ad ]] && \
Expand Down Expand Up @@ -142,7 +142,7 @@ import groovy.json.JsonSlurper
mitoFile=${outdir}/${sample_info.species}.MitoCarta2.0.txt
python $projectDir/bin/stPreprocess.py --filePath=\${dname}/ --npFactorsOutputName=st_adata_counts_in_tissue_factors.npz --rawAdata=st_adata_raw.h5ad --mitoFile=\$mitoFile --pltFigSize=params.STpreprocess_pltFigSize --minCounts=params.STpreprocess_minCounts --minGenes=params.STpreprocess_minGenes --minCells=params.STpreprocess_minCells --histplotQCmaxTotalCounts=params.STpreprocess_histplotQCmaxTotalCounts --histplotQCminGeneCounts=params.STpreprocess_histplotQCminGeneCounts --histplotQCbins=params.STpreprocess_histplotQCbins
python $projectDir/bin/stPreprocess.py --filePath=\${dname}/ --npFactorsOutputName=st_adata_counts_in_tissue_factors.npz --rawAdata=st_adata_raw.h5ad --mitoFile=\$mitoFile --pltFigSize=$params.STpreprocess_pltFigSize --minCounts=$params.STpreprocess_minCounts --minGenes=$params.STpreprocess_minGenes --minCells=$params.STpreprocess_minCells --histplotQCmaxTotalCounts=$params.STpreprocess_histplotQCmaxTotalCounts --histplotQCminGeneCounts=$params.STpreprocess_histplotQCminGeneCounts --histplotQCbins=$params.STpreprocess_histplotQCbins
if [[ -s \${dname}/st_adata_norm.h5ad ]] && \
[[ -s \${dname}/st_adata_X.npz ]] && \
Expand Down Expand Up @@ -185,7 +185,7 @@ import groovy.json.JsonSlurper
mitoFile=${outdir}/${sample_info.species}.MitoCarta2.0.txt
python $projectDir/bin/scPreprocess.py --filePath=\${dname}/ --npFactorsOutputName=sc_adata_counts_factors.npz --rawAdata=sc_adata_raw.h5ad --mitoFile=\$mitoFile --pltFigSize=params.SCpreprocess_pltFigSize --minCounts=params.SCpreprocess_minCounts --minGenes=params.SCpreprocess_minGenes --minCells=params.SCpreprocess_minCells --histplotQCmaxTotalCounts=params.SCpreprocess_histplotQCmaxTotalCounts --histplotQCminGeneCounts=params.SCpreprocess_histplotQCminGeneCounts --histplotQCbins=params.SCpreprocess_histplotQCbins
python $projectDir/bin/scPreprocess.py --filePath=\${dname}/ --npFactorsOutputName=sc_adata_counts_factors.npz --rawAdata=sc_adata_raw.h5ad --mitoFile=\$mitoFile --pltFigSize=$params.SCpreprocess_pltFigSize --minCounts=$params.SCpreprocess_minCounts --minGenes=$params.SCpreprocess_minGenes --minCells=$params.SCpreprocess_minCells --histplotQCmaxTotalCounts=$params.SCpreprocess_histplotQCmaxTotalCounts --histplotQCminGeneCounts=$params.SCpreprocess_histplotQCminGeneCounts --histplotQCbins=$params.SCpreprocess_histplotQCbins
if [[ -s \${dname}/sc_adata_norm.h5ad ]] && \
[[ -s \${dname}/sc_adata_X.npz ]] && \
Expand Down Expand Up @@ -237,7 +237,7 @@ import groovy.json.JsonSlurper
dname=${outdir}/\${sample_id}
Rscript $projectDir/bin/characterization_STdeconvolve.R --filePath=\${dname}/ --outsPath=${sample_info.st_data_dir} --mtxGeneColumn=params.STdeconvolve_mtxGeneColumn --countsFactor=params.STdeconvolve_countsFactor --corpusRemoveAbove=params.STdeconvolve_corpusRemoveAbove --corpusRemoveBelow=params.STdeconvolve_corpusRemoveBelow --LDAminTopics=params.STdeconvolve_LDAminTopics --LDAmaxTopics=params.STdeconvolve_LDAmaxTopics --STdeconvolveScatterpiesSize=params.STdeconvolve_ScatterpiesSize --STdeconvolveFeaturesSizeFactor=params.STdeconvolve_FeaturesSizeFactor
Rscript $projectDir/bin/characterization_STdeconvolve.R --filePath=\${dname}/ --outsPath=${sample_info.st_data_dir} --mtxGeneColumn=$params.STdeconvolve_mtxGeneColumn --countsFactor=$params.STdeconvolve_countsFactor --corpusRemoveAbove=$params.STdeconvolve_corpusRemoveAbove --corpusRemoveBelow=$params.STdeconvolve_corpusRemoveBelow --LDAminTopics=$params.STdeconvolve_LDAminTopics --LDAmaxTopics=$params.STdeconvolve_LDAmaxTopics --STdeconvolveScatterpiesSize=$params.STdeconvolve_ScatterpiesSize --STdeconvolveFeaturesSizeFactor=$params.STdeconvolve_FeaturesSizeFactor
if [[ -s \${dname}/STdeconvolve_prop_norm.csv ]] && \
[[ -s \${dname}/STdeconvolve_beta_norm.csv ]] && \
Expand Down Expand Up @@ -281,7 +281,7 @@ import groovy.json.JsonSlurper
dname=${outdir}/\${sample_id}
Rscript $projectDir/bin/characterization_SPOTlight.R --filePath=\${dname}/ --outsPath=${sample_info.st_data_dir} --mtxGeneColumn=params.SPOTlight_mtxGeneColumn --countsFactor=params.SPOTlight_countsFactor --clusterResolution=params.SPOTlight_clusterResolution --numberHVG=params.SPOTlight_numberHVG --numberCellsPerCelltype=params.SPOTlight_numberCellsPerCelltype --SPOTlightScatterpiesSize=params.SPOTlight_ScatterpiesSize
Rscript $projectDir/bin/characterization_SPOTlight.R --filePath=\${dname}/ --outsPath=${sample_info.st_data_dir} --mtxGeneColumn=$params.SPOTlight_mtxGeneColumn --countsFactor=$params.SPOTlight_countsFactor --clusterResolution=$params.SPOTlight_clusterResolution --numberHVG=$params.SPOTlight_numberHVG --numberCellsPerCelltype=$params.SPOTlight_numberCellsPerCelltype --SPOTlightScatterpiesSize=$params.SPOTlight_ScatterpiesSize
if [[ -s \${dname}/SPOTlight_prop_norm.csv ]] && \
[[ -s \${dname}/SPOTlight_beta_norm.csv ]] && \
Expand Down Expand Up @@ -325,7 +325,7 @@ import groovy.json.JsonSlurper
dname=${outdir}/\${sample_id}
Rscript $projectDir/bin/characterization_BayesSpace.R --filePath=\${dname}/ --numberHVG=params.BayesSpace_numberHVG --numberPCs=params.BayesSpace_numberPCs --minClusters=params.BayesSpace_minClusters --maxClusters=params.BayesSpace_maxClusters --optimalQ=params.BayesSpace_optimalQ --STplatform=params.BayesSpace_STplatform
Rscript $projectDir/bin/characterization_BayesSpace.R --filePath=\${dname}/ --numberHVG=$params.BayesSpace_numberHVG --numberPCs=$params.BayesSpace_numberPCs --minClusters=$params.BayesSpace_minClusters --maxClusters=$params.BayesSpace_maxClusters --optimalQ=$params.BayesSpace_optimalQ --STplatform=$params.BayesSpace_STplatform
if [[ -s \${dname}/bayes_spot_cluster.csv ]] && \
[[ -s \${dname}/bayes_subspot_cluster_and_coord.csv ]] && \
Expand Down Expand Up @@ -366,7 +366,7 @@ import groovy.json.JsonSlurper
dname=${outdir}/\${sample_id}
python $projectDir/bin/stSpatialDE.py --filePath=\${dname}/ --numberOfColumns=params.SpatialDE_numberOfColumns
python $projectDir/bin/stSpatialDE.py --filePath=\${dname}/ --numberOfColumns=$params.SpatialDE_numberOfColumns
if [[ -s \${dname}/stSpatialDE.csv ]]
then
Expand Down Expand Up @@ -419,7 +419,7 @@ import groovy.json.JsonSlurper
dname=${outdir}/\${sample_id}
python $projectDir/bin/stClusteringWorkflow.py --filePath=\${dname}/ --resolution=params.Clustering_resolution
python $projectDir/bin/stClusteringWorkflow.py --filePath=\${dname}/ --resolution=$params.Clustering_resolution
if [[ -s \${dname}/st_adata_processed.h5ad ]] && \
[[ -s \${dname}/sc_adata_processed.h5ad ]]
Expand Down

0 comments on commit e45e1d6

Please sign in to comment.