library(imageFeatureTCGA)
library(HistoImagePlot)
library(dplyr)
library(SpatialExperiment)
library(ggplot2)
if (!requireNamespace("BiocManager", quietly = TRUE))
install.packages("BiocManager")
BiocManager::install("waldronlab/HistoImagePlot")
HoverNet is a deep learning model for simultaneous segmentation and classification of nuclei in multi-tissue histology images. This vignette demonstrates how to import HoverNet output files into Bioconductor’s spatial data structures and create visualizations of the segmentation results.
The package supports two main file formats:
.json or .json.gz): Contains cell coordinates,
types, and optional contours.h5ad): AnnData format with additional computed
features like mean intensity and nearest neighbor distanceSpatialExperiment
object from URLs (with Automatic Caching) The file will be automatically
cached using BiocFileCache:
hov_file <- paste0(
"https://store.cancerdatasci.org/hovernet/h5ad/",
"TCGA-23-1021-01Z-00-DX1.F07C221B-D401-47A5-9519-10DE59CA1E9D.h5ad.gz")
thumb_path <- paste0(
"https://store.cancerdatasci.org/hovernet/thumb/",
"TCGA-23-1021-01Z-00-DX1.F07C221B-D401-47A5-9519-10DE59CA1E9D.png")
hn_spe <- HoverNet(hov_file, outClass = "SpatialExperiment") |>
import()
The package provides a convenient function to overlay the segmentation on the original tissue thumbnail image.
thumb_path <- paste0(
"https://store.cancerdatasci.org/hovernet/thumb/",
"TCGA-23-1021-01Z-00-DX1.F07C221B-D401-47A5-9519-10DE59CA1E9D.png")
plotHoverNetH5ADOverlay(hn_spe, thumb_path)
plotHoverNetH5ADOverlay(
hn_spe,
thumb_path,
title = "Ovarian Cancer Tissue - Cell Segmentation",
point_size = 0.02,
legend_point_size = 3
)
custom_colors <- c(
"no label" = "#808080",
"neoplastic" = "#E31A1C",
"inflammatory" = "#1F78B4",
"stromal" = "#33A02C",
"necrotic" = "#FF7F00",
"benign epithelial" = "#6A3D9A"
)
plotHoverNetH5ADOverlay(
hn_spe,
thumb_path,
color_palette = custom_colors,
title = "Custom Color Scheme"
)
H5AD files contain additional computed features like mean intensity and nearest neighbor distance.
h5ad_coords <- data.frame(spatialCoords(hn_spe), colData(hn_spe))
p1 <- ggplot(h5ad_coords, aes(x = x_centroid, y = y_centroid,
color = mean_intensity)) +
geom_point(size = 0.5) +
scale_color_viridis_c() +
coord_fixed() +
theme_minimal() +
labs(title = "Mean Intensity", color = "Intensity")
p2 <- ggplot(h5ad_coords, aes(x = x_centroid, y = y_centroid,
color = nearest_neighbor_distance)) +
geom_point(size = 0.5) +
scale_color_viridis_c(option = "plasma") +
coord_fixed() +
theme_minimal() +
labs(title = "Nearest Neighbor Distance", color = "Distance")
cowplot::plot_grid(p1, p2, ncol = 2)
Click here for Session Info
sessionInfo()
#> R Under development (unstable) (2025-10-28 r88973)
#> Platform: x86_64-pc-linux-gnu
#> Running under: Ubuntu 24.04.4 LTS
#>
#> Matrix products: default
#> BLAS/LAPACK: /usr/lib/x86_64-linux-gnu/openblas-pthread/libopenblasp-r0.3.26.so; LAPACK version 3.12.0
#>
#> locale:
#> [1] LC_CTYPE=en_US.UTF-8 LC_NUMERIC=C 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 LC_PAPER=en_US.UTF-8 LC_NAME=C
#> [9] LC_ADDRESS=C LC_TELEPHONE=C LC_MEASUREMENT=en_US.UTF-8 LC_IDENTIFICATION=C
#>
#> time zone: America/New_York
#> tzcode source: system (glibc)
#>
#> attached base packages:
#> [1] stats4 stats graphics grDevices utils datasets methods base
#>
#> other attached packages:
#> [1] ggplot2_4.0.1 SpatialExperiment_1.21.0 SingleCellExperiment_1.33.0 SummarizedExperiment_1.41.0
#> [5] Biobase_2.71.0 GenomicRanges_1.63.1 Seqinfo_1.1.0 IRanges_2.45.0
#> [9] S4Vectors_0.49.0 BiocGenerics_0.57.0 generics_0.1.4 MatrixGenerics_1.23.0
#> [13] matrixStats_1.5.0 dplyr_1.1.4 HistoImagePlot_0.99.7 imageFeatureTCGA_0.99.56
#> [17] colorout_1.3-2
#>
#> loaded via a namespace (and not attached):
#> [1] DBI_1.2.3 httr2_1.2.2 remotes_2.5.0 anndataR_1.1.0 rlang_1.1.6 magrittr_2.0.4
#> [7] otel_0.2.0 compiler_4.6.0 RSQLite_2.4.5 png_0.1-8 vctrs_0.6.5 pkgconfig_2.0.3
#> [13] fastmap_1.2.0 dbplyr_2.5.1 magick_2.9.0 XVector_0.51.0 ellipsis_0.3.2 labeling_0.4.3
#> [19] rmarkdown_2.30 sessioninfo_1.2.3 tzdb_0.5.0 purrr_1.2.0 bit_4.6.0 xfun_0.56
#> [25] gert_2.3.1 cachem_1.1.0 jsonlite_2.0.0 blob_1.2.4 rhdf5filters_1.23.3 DelayedArray_0.37.0
#> [31] Rhdf5lib_1.33.0 R6_2.6.1 RColorBrewer_1.1-3 reticulate_1.44.1 pkgload_1.4.1 Rcpp_1.1.1
#> [37] knitr_1.51 usethis_3.2.1 readr_2.1.6 BiocBaseUtils_1.13.0 Matrix_1.7-4 tidyselect_1.2.1
#> [43] rstudioapi_0.18.0 dichromat_2.0-0.1 abind_1.4-8 yaml_2.3.12 codetools_0.2-20 curl_7.0.0
#> [49] rjsoncons_1.3.2 pkgbuild_1.4.8 lattice_0.22-7 tibble_3.3.0 withr_3.0.2 S7_0.2.1
#> [55] askpass_1.2.1 evaluate_1.0.5 BiocAddins_0.99.26 desc_1.4.3 BiocFileCache_3.1.0 pillar_1.11.1
#> [61] BiocManager_1.30.27 filelock_1.0.3 rsconnect_1.7.0 rprojroot_2.1.1 credentials_2.0.3 hms_1.1.4
#> [67] scales_1.4.0 glue_1.8.0 tools_4.6.0 BiocIO_1.21.0 sys_3.4.3 fs_1.6.6
#> [73] cowplot_1.2.0 rhdf5_2.55.12 grid_4.6.0 devtools_2.4.6 TENxIO_1.13.3 cli_3.6.5
#> [79] rappdirs_0.3.4 S4Arrays_1.11.1 viridisLite_0.4.2 gtable_0.3.6 digest_0.6.39 SparseArray_1.11.10
#> [85] rjson_0.2.23 farver_2.1.2 memoise_2.0.1 htmltools_0.5.9 lifecycle_1.0.4 openssl_2.3.4
#> [91] bit64_4.6.0-1