This vignette shows how to build a SpatialExperiment
(SPE) from: - Nanostring CosMx (RNA/protein) outputs -
10x Genomics Xenium outputs
For each technology, we demonstrate: - Route A:
SpaceTrooper’s high-level reader (readCosmxSPE,
readCosmxProteinSPE, readXeniumSPE) -
Route B: read with
SpatialExperimentIO, then standardize with
updateCosmxSPE / updateXeniumSPE
We begin with CosMx. The package ships with a small CosMx example for demonstration.
cospath <- system.file(file.path("extdata", "CosMx_DBKero_Tiny"), package="SpaceTrooper")
cospath
#> [1] "/tmp/RtmpkQaa8M/Rinst200651542705/SpaceTrooper/extdata/CosMx_DBKero_Tiny"Use readCosmxSPE() to construct an SPE from CosMx
outputs; it also normalizes names/metadata and records polygons/FOV info
if present.
spe_cos <- readCosmxSPE(
dirName=cospath,
sampleName="DBKero_Tiny",
coordNames=c("CenterX_global_px", "CenterY_global_px"),
countMatFPattern="exprMat_file.csv",
metadataFPattern="metadata_file.csv",
polygonsFPattern="polygons.csv",
fovPosFPattern="fov_positions_file.csv",
fovdims=c(xdim=4256, ydim=4256)
)
spe_cos
#> class: SpatialExperiment
#> dim: 1010 905
#> metadata(4): fov_positions fov_dim polygons technology
#> assays(1): counts
#> rownames(1010): RAMP2 CD83 ... NegPrb09 NegPrb10
#> rowData names(0):
#> colnames(905): f16_c1 f16_c10 ... f16_c98 f16_c99
#> colData names(20): fov cellID ... sample_id cell_id
#> reducedDimNames(0):
#> mainExpName: NULL
#> altExpNames(0):
#> spatialCoords names(2) : CenterX_global_px CenterY_global_px
#> imgData names(1): sample_idInspect essentials:
assayNames(spe_cos)
#> [1] "counts"
dim(spe_cos)
#> [1] 1010 905
head(colnames(spatialCoords(spe_cos)))
#> [1] "CenterX_global_px" "CenterY_global_px"
metadata(spe_cos)$technology
#> [1] "Nanostring_CosMx"
metadata(spe_cos)$polygons
#> [1] "/tmp/RtmpkQaa8M/Rinst200651542705/SpaceTrooper/extdata/CosMx_DBKero_Tiny/DBKero-polygons.csv"If working with CosMx Protein, use the convenience wrapper:
protfolder <- system.file("extdata", "S01_prot", package = "SpaceTrooper")
spe_cos_prot <- readCosmxProteinSPE(
dirName=protfolder,
sampleName="cosmx_prots",
coordNames=c("CenterX_global_px", "CenterY_global_px"),
countMatFPattern="exprMat_file.csv",
metadataFPattern="metadata_file.csv",
polygonsFPattern="polygons.csv",
fovPosFPattern="fov_positions_file.csv",
fovdims=c(xdim=4256, ydim=4256)
)
metadata(spe_cos_prot)$technologyIf you prefer to read with SpatialExperimentIO
first, upgrade the object with updateCosmxSPE() to
harmonize names/metadata and attach polygons.
spe_cos_raw <- SpatialExperimentIO::readCosmxSXE(
dirName=cospath,
returnType="SPE",
countMatPattern="exprMat_file.csv",
metaDataPattern="metadata_file.csv",
coordNames=c("CenterX_global_px", "CenterY_global_px"),
addFovPos=TRUE,
fovPosPattern="fov_positions_file.csv",
altExps=NULL,
addParquetPaths=FALSE
)
spe_cos_std <- updateCosmxSPE(
spe=spe_cos_raw,
dirName=cospath,
sampleName="DBKero_Tiny",
polygonsFPattern="polygons.csv",
fovdims=c(xdim=4256, ydim=4256)
)
identical(spe_cos_std, spe_cos)
#> [1] TRUEA small Xenium example is also included for demonstration.
xepath <- system.file("extdata", "Xenium_small", package = "SpaceTrooper")
xepath
#> [1] "/tmp/RtmpkQaa8M/Rinst200651542705/SpaceTrooper/extdata/Xenium_small"readXeniumSPE() builds the SPE from a Xenium Output
Bundle (root or outs/) and standardizes metadata.
Key options: - type: "HDF5" or
"sparse" (feature matrix) - boundariesType:
"parquet" or "csv" (cell boundaries) -
computeMissingMetrics: compute QC metrics (area/aspect
ratio) if needed - keepPolygons: append polygons to
colData - addFOVs: derive FOV IDs from
transcript parquet
spe_xen_a <- readXeniumSPE(
dirName=xepath,
type="HDF5",
coordNames=c("x_centroid", "y_centroid"),
boundariesType="parquet",
computeMissingMetrics=TRUE,
keepPolygons=TRUE,
countsFilePattern="cell_feature_matrix",
metadataFPattern="cells.csv.gz",
polygonsFPattern="cell_boundaries",
polygonsCol="polygons",
txPattern="transcripts",
addFOVs=FALSE
)
#> Computing missing metrics, this could take some time...
spe_xen_a
#> class: SpatialExperiment
#> dim: 4 6
#> metadata(2): polygons technology
#> assays(1): counts
#> rownames(4): ABCC11 ACTA2 ACTG2 ADAM9
#> rowData names(3): ID Symbol Type
#> colnames(6): 1 2 ... 5 6
#> colData names(11): X cell_id ... Area_um polygons
#> reducedDimNames(0):
#> mainExpName: NULL
#> altExpNames(0):
#> spatialCoords names(2) : x_centroid y_centroid
#> imgData names(1): sample_idQuick checks:
assayNames(spe_xen_a)
#> [1] "counts"
dim(spe_xen_a)
#> [1] 4 6
colnames(spatialCoords(spe_xen_a))
#> [1] "x_centroid" "y_centroid"
metadata(spe_xen_a)$polygons
#> [1] "/tmp/RtmpkQaa8M/Rinst200651542705/SpaceTrooper/extdata/Xenium_small/cell_boundaries.parquet"
metadata(spe_xen_a)$technology
#> [1] "10X_Xenium"Read with SpatialExperimentIO and then pass through
updateXeniumSPE() for SpaceTrooper-standardized metadata
and optional metrics/FOV extraction.
spe_xen_b <- SpatialExperimentIO::readXeniumSXE(
dirName=xepath,
countMatPattern="cell_feature_matrix.h5",
metaDataPattern="cells.csv.gz",
coordNames=c("x_centroid", "y_centroid"),
returnType="SPE",
addExperimentXenium=FALSE,
altExps=NULL,
addParquetPaths=FALSE
)
spe_xen_b <- updateXeniumSPE(
spe=spe_xen_b,
dirName=xepath,
boundariesType="parquet",
computeMissingMetrics=TRUE,
keepPolygons=TRUE,
polygonsFPattern="cell_boundaries",
polygonsCol="polygons",
txPattern="transcripts",
addFOVs=FALSE
)
#> Computing missing metrics, this could take some time...
spe_xen_b
#> class: SpatialExperiment
#> dim: 4 6
#> metadata(2): polygons technology
#> assays(1): counts
#> rownames(4): ABCC11 ACTA2 ACTG2 ADAM9
#> rowData names(3): ID Symbol Type
#> colnames(6): 1 2 ... 5 6
#> colData names(11): X cell_id ... Area_um polygons
#> reducedDimNames(0):
#> mainExpName: NULL
#> altExpNames(0):
#> spatialCoords names(2) : x_centroid y_centroid
#> imgData names(1): sample_idValidate:
identical(metadata(spe_xen_b)$technology, "10X_Xenium")
#> [1] TRUE
identical(spe_xen_a, spe_xen_b)
#> [1] TRUEmetadata_file.csv), expression matrix
(exprMat_file.csv), optional polygon CSV
(polygons.csv), and FOV positions.
updateCosmxSPE() also fixes common field names and records
FOV dims in metadata.outs/ folder. Feature matrix may be
cell_feature_matrix.h5 (HDF5) or sparse folder; cells
metadata cells.csv.gz; boundaries as .parquet
or .csv.gz; transcript parquet for FOV attribution.
readXeniumSPE() auto-detects outs/ if you pass
the bundle root.sessionInfo()
#> R version 4.6.1 (2026-06-24)
#> Platform: x86_64-pc-linux-gnu
#> Running under: Ubuntu 26.04 LTS
#>
#> Matrix products: default
#> BLAS: /usr/lib/x86_64-linux-gnu/openblas-pthread/libblas.so.3
#> LAPACK: /usr/lib/x86_64-linux-gnu/openblas-pthread/libopenblasp-r0.3.32.so; LAPACK version 3.12.0
#>
#> locale:
#> [1] LC_CTYPE=en_US.UTF-8 LC_NUMERIC=C
#> [3] LC_TIME=en_US.UTF-8 LC_COLLATE=en_US.UTF-8
#> [5] LC_MONETARY=en_US.UTF-8 LC_MESSAGES=en_US.UTF-8
#> [7] LC_PAPER=en_US.UTF-8 LC_NAME=C
#> [9] LC_ADDRESS=C LC_TELEPHONE=C
#> [11] LC_MEASUREMENT=en_US.UTF-8 LC_IDENTIFICATION=C
#>
#> time zone: Etc/UTC
#> tzcode source: system (glibc)
#>
#> attached base packages:
#> [1] stats4 stats graphics grDevices utils datasets methods
#> [8] base
#>
#> other attached packages:
#> [1] SpaceTrooper_1.3.0 SpatialExperiment_1.23.0
#> [3] SingleCellExperiment_1.35.2 SummarizedExperiment_1.43.0
#> [5] Biobase_2.73.1 GenomicRanges_1.65.1
#> [7] Seqinfo_1.3.0 IRanges_2.47.2
#> [9] S4Vectors_0.51.5 BiocGenerics_0.59.10
#> [11] generics_0.1.4 MatrixGenerics_1.25.0
#> [13] matrixStats_1.5.0 BiocStyle_2.41.0
#>
#> loaded via a namespace (and not attached):
#> [1] DBI_1.3.0 gridExtra_2.3.1
#> [3] rlang_1.3.0 magrittr_2.0.5
#> [5] scater_1.41.2 otel_0.2.0
#> [7] e1071_1.7-17 compiler_4.6.1
#> [9] DelayedMatrixStats_1.35.0 SpatialExperimentIO_1.5.0
#> [11] sfheaders_0.4.5 vctrs_0.7.3
#> [13] shape_1.4.6.1 pkgconfig_2.0.3
#> [15] fastmap_1.2.0 backports_1.5.1
#> [17] magick_2.9.1 XVector_0.53.0
#> [19] scuttle_1.23.1 rmarkdown_2.31
#> [21] ggbeeswarm_0.7.3 purrr_1.2.2
#> [23] bit_4.6.0 glmnet_5.0
#> [25] xfun_0.60 cachem_1.1.0
#> [27] beachmat_2.29.0 jsonlite_2.0.0
#> [29] rhdf5filters_1.25.0 DelayedArray_0.39.3
#> [31] Rhdf5lib_2.1.0 BiocParallel_1.47.0
#> [33] irlba_2.3.7 broom_1.0.13
#> [35] parallel_4.6.1 R6_2.6.1
#> [37] bslib_0.11.0 RColorBrewer_1.1-3
#> [39] limma_3.69.2 car_3.1-5
#> [41] jquerylib_0.1.4 iterators_1.0.14
#> [43] Rcpp_1.1.2 assertthat_0.2.1
#> [45] knitr_1.51 R.utils_2.13.0
#> [47] splines_4.6.1 Matrix_1.7-5
#> [49] tidyselect_1.2.1 viridis_0.6.5
#> [51] abind_1.4-8 yaml_2.3.12
#> [53] codetools_0.2-20 lattice_0.22-9
#> [55] tibble_3.3.1 S7_0.2.2
#> [57] evaluate_1.0.5 sf_1.1-1
#> [59] survival_3.8-9 units_1.0-1
#> [61] proxy_0.4-29 pillar_1.11.1
#> [63] BiocManager_1.30.27 ggpubr_1.0.0
#> [65] carData_3.0-6 KernSmooth_2.23-26
#> [67] foreach_1.5.2 ggplot2_4.0.3
#> [69] sparseMatrixStats_1.25.0 scales_1.4.0
#> [71] class_7.3-23 glue_1.8.1
#> [73] maketools_1.3.2 tools_4.6.1
#> [75] BiocNeighbors_2.7.2 robustbase_0.99-7
#> [77] sys_3.4.3 data.table_1.18.4
#> [79] ScaledMatrix_1.21.0 locfit_1.5-9.12
#> [81] ggsignif_0.6.4 buildtools_1.0.0
#> [83] cowplot_1.2.0 rhdf5_2.57.1
#> [85] grid_4.6.1 tidyr_1.3.2
#> [87] DropletUtils_1.33.0 edgeR_4.11.4
#> [89] beeswarm_0.4.0 BiocSingular_1.29.0
#> [91] HDF5Array_1.41.0 vipor_0.4.7
#> [93] rsvd_1.0.5 Formula_1.2-5
#> [95] cli_3.6.6 viridisLite_0.4.3
#> [97] S4Arrays_1.13.0 arrow_25.0.0
#> [99] dplyr_1.2.1 DEoptimR_1.2-0
#> [101] gtable_0.3.6 R.methodsS3_1.8.2
#> [103] rstatix_1.0.0 sass_0.4.10
#> [105] digest_0.6.39 classInt_0.4-11
#> [107] ggrepel_0.9.8 SparseArray_1.13.2
#> [109] dqrng_0.4.1 rjson_0.2.23
#> [111] farver_2.1.2 htmltools_0.5.9
#> [113] R.oo_1.27.1 lifecycle_1.0.5
#> [115] h5mread_1.5.0 statmod_1.5.2
#> [117] bit64_4.8.2