Downsampling
Often, high-dimensional cytometry experiments collect tens or hundreds or millions of cells in total, and it can be useful to downsample to a smaller, more computationally tractable number of cells - either for a final analysis or while developing code.
To do this, {tidytof} implements the
tof_downsample() verb, which allows downsampling using 3
methods: downsampling to an integer number of cells, downsampling to a
fixed proportion of the total number of input cells, or downsampling to
a fixed cellular density in phenotypic space.
Downsampling with tof_downsample()
Using {tidytof}’s built-in dataset
phenograph_data, we can see that the original size of the
dataset is 1000 cells per cluster, or 3000 cells in total:
data(phenograph_data)
phenograph_data |>
dplyr::count(phenograph_cluster)
#> # A tibble: 3 × 2
#> phenograph_cluster n
#> <chr> <int>
#> 1 cluster1 1000
#> 2 cluster2 1000
#> 3 cluster3 1000To randomly sample 200 cells per cluster, we can use
tof_downsample() using the “constant”
method:
phenograph_data |>
# downsample
tof_downsample(
group_cols = phenograph_cluster,
method = "constant",
num_cells = 200
) |>
# count the number of downsampled cells in each cluster
count(phenograph_cluster)
#> # A tibble: 3 × 2
#> phenograph_cluster n
#> <chr> <int>
#> 1 cluster1 200
#> 2 cluster2 200
#> 3 cluster3 200Alternatively, if we wanted to sample 50% of the cells in each
cluster, we could use the “prop” method:
phenograph_data |>
# downsample
tof_downsample(
group_cols = phenograph_cluster,
method = "prop",
prop_cells = 0.5
) |>
# count the number of downsampled cells in each cluster
count(phenograph_cluster)
#> # A tibble: 3 × 2
#> phenograph_cluster n
#> <chr> <int>
#> 1 cluster1 500
#> 2 cluster2 500
#> 3 cluster3 500And finally, we might also be interested in taking a slightly
different approach to downsampling that reduces the number of cells not
to a fixed constant or proportion, but to a fixed density in
phenotypic space. For example, the following scatterplot demonstrates
that there are certain areas of phenotypic density in
phenograph_data that contain more cells than others along
the cd34/cd38 axes:
rescale_max <-
function(x, to = c(0, 1), from = range(x, na.rm = TRUE)) {
x / from[2] * to[2]
}
phenograph_data |>
# preprocess all numeric columns in the dataset
tof_preprocess(undo_noise = FALSE) |>
# plot
ggplot(aes(x = cd34, y = cd38)) +
geom_hex() +
coord_fixed(ratio = 0.4) +
scale_x_continuous(limits = c(NA, 1.5)) +
scale_y_continuous(limits = c(NA, 4)) +
scale_fill_viridis_c(
labels = function(x) round(rescale_max(x), 2)
) +
labs(
fill = "relative density"
)
To reduce the number of cells in our dataset until the local density
around each cell in our dataset is relatively constant, we can use the
“density” method of tof_downsample:
phenograph_data |>
tof_preprocess(undo_noise = FALSE) |>
tof_downsample(method = "density", density_cols = c(cd34, cd38)) |>
# plot
ggplot(aes(x = cd34, y = cd38)) +
geom_hex() +
coord_fixed(ratio = 0.4) +
scale_x_continuous(limits = c(NA, 1.5)) +
scale_y_continuous(limits = c(NA, 4)) +
scale_fill_viridis_c(
labels = function(x) round(rescale_max(x), 2)
) +
labs(
fill = "relative density"
)
Thus, we can see that the density after downsampling is more uniform
(though not exactly uniform) across the range of
cd34/cd38 values in
phenograph_data.
Additional documentation
For more details, check out the documentation for the 3 underlying
members of the tof_downsample_* function family (which are
wrapped by tof_downsample):
tof_downsample_constanttof_downsample_proptof_downsample_density
Session info
sessionInfo()
#> R version 4.4.2 (2024-10-31)
#> Platform: x86_64-pc-linux-gnu
#> Running under: Ubuntu 24.04.1 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-r0.3.26.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=C
#> [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] tidyr_1.3.1 stringr_1.5.1
#> [3] HDCytoData_1.26.0 flowCore_2.19.0
#> [5] SummarizedExperiment_1.37.0 Biobase_2.67.0
#> [7] GenomicRanges_1.59.1 GenomeInfoDb_1.43.2
#> [9] IRanges_2.41.2 S4Vectors_0.45.2
#> [11] MatrixGenerics_1.19.1 matrixStats_1.5.0
#> [13] ExperimentHub_2.15.0 AnnotationHub_3.15.0
#> [15] BiocFileCache_2.15.0 dbplyr_2.5.0
#> [17] BiocGenerics_0.53.3 generics_0.1.3
#> [19] forcats_1.0.0 ggplot2_3.5.1
#> [21] dplyr_1.1.4 tidytof_1.1.0
#> [23] rmarkdown_2.29
#>
#> loaded via a namespace (and not attached):
#> [1] sys_3.4.3 jsonlite_1.8.9 shape_1.4.6.1
#> [4] magrittr_2.0.3 farver_2.1.2 vctrs_0.6.5
#> [7] memoise_2.0.1 htmltools_0.5.8.1 S4Arrays_1.7.1
#> [10] curl_6.1.0 SparseArray_1.7.3 sass_0.4.9
#> [13] parallelly_1.41.0 bslib_0.8.0 lubridate_1.9.4
#> [16] cachem_1.1.0 buildtools_1.0.0 igraph_2.1.3
#> [19] mime_0.12 lifecycle_1.0.4 iterators_1.0.14
#> [22] pkgconfig_2.0.3 Matrix_1.7-1 R6_2.5.1
#> [25] fastmap_1.2.0 GenomeInfoDbData_1.2.13 future_1.34.0
#> [28] digest_0.6.37 colorspace_2.1-1 AnnotationDbi_1.69.0
#> [31] irlba_2.3.5.1 RSQLite_2.3.9 labeling_0.4.3
#> [34] filelock_1.0.3 cytolib_2.19.1 yardstick_1.3.1
#> [37] timechange_0.3.0 httr_1.4.7 polyclip_1.10-7
#> [40] abind_1.4-8 compiler_4.4.2 bit64_4.5.2
#> [43] withr_3.0.2 doParallel_1.0.17 viridis_0.6.5
#> [46] DBI_1.2.3 ggforce_0.4.2 MASS_7.3-64
#> [49] lava_1.8.1 embed_1.1.4 rappdirs_0.3.3
#> [52] DelayedArray_0.33.3 tools_4.4.2 future.apply_1.11.3
#> [55] nnet_7.3-20 glue_1.8.0 grid_4.4.2
#> [58] Rtsne_0.17 recipes_1.1.0 gtable_0.3.6
#> [61] tzdb_0.4.0 class_7.3-23 data.table_1.16.4
#> [64] hms_1.1.3 utf8_1.2.4 tidygraph_1.3.1
#> [67] XVector_0.47.2 RcppAnnoy_0.0.22 ggrepel_0.9.6
#> [70] BiocVersion_3.21.1 foreach_1.5.2 pillar_1.10.1
#> [73] RcppHNSW_0.6.0 splines_4.4.2 tweenr_2.0.3
#> [76] lattice_0.22-6 survival_3.8-3 bit_4.5.0.1
#> [79] RProtoBufLib_2.19.0 tidyselect_1.2.1 Biostrings_2.75.3
#> [82] maketools_1.3.1 knitr_1.49 gridExtra_2.3
#> [85] xfun_0.50 graphlayouts_1.2.1 hardhat_1.4.0
#> [88] timeDate_4041.110 stringi_1.8.4 UCSC.utils_1.3.1
#> [91] yaml_2.3.10 evaluate_1.0.3 codetools_0.2-20
#> [94] ggraph_2.2.1 tibble_3.2.1 BiocManager_1.30.25
#> [97] cli_3.6.3 uwot_0.2.2 rpart_4.1.24
#> [100] munsell_0.5.1 jquerylib_0.1.4 Rcpp_1.0.14
#> [103] globals_0.16.3 png_0.1-8 parallel_4.4.2
#> [106] gower_1.0.2 readr_2.1.5 blob_1.2.4
#> [109] listenv_0.9.1 glmnet_4.1-8 viridisLite_0.4.2
#> [112] ipred_0.9-15 ggridges_0.5.6 scales_1.3.0
#> [115] prodlim_2024.06.25 purrr_1.0.2 crayon_1.5.3
#> [118] rlang_1.1.4 KEGGREST_1.47.0