tests/testthat/test_measureObjects.R
In BodenmillerGroup/SingleCellMapper: Visualization of highly multiplexed imaging data in R
test_that("measureObjects: defaults work", {
data("pancreasSCE")
data("pancreasMasks")
data("pancreasImages")
# Works
expect_silent(sce <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb"))
cur_H3 <- c(tapply(pancreasImages[[1]][,,1], pancreasMasks[[1]], mean)[-1],
tapply(pancreasImages[[2]][,,1], pancreasMasks[[2]], mean)[-1],
tapply(pancreasImages[[3]][,,1], pancreasMasks[[3]], mean)[-1])
cur_CD99 <- c(tapply(pancreasImages[[1]][,,2], pancreasMasks[[1]], mean)[-1],
tapply(pancreasImages[[2]][,,2], pancreasMasks[[2]], mean)[-1],
tapply(pancreasImages[[3]][,,2], pancreasMasks[[3]], mean)[-1])
cur_PIN <- c(tapply(pancreasImages[[1]][,,3], pancreasMasks[[1]], mean)[-1],
tapply(pancreasImages[[2]][,,3], pancreasMasks[[2]], mean)[-1],
tapply(pancreasImages[[3]][,,3], pancreasMasks[[3]], mean)[-1])
cur_CD8a <- c(tapply(pancreasImages[[1]][,,4], pancreasMasks[[1]], mean)[-1],
tapply(pancreasImages[[2]][,,4], pancreasMasks[[2]], mean)[-1],
tapply(pancreasImages[[3]][,,4], pancreasMasks[[3]], mean)[-1])
cur_CDH <- c(tapply(pancreasImages[[1]][,,5], pancreasMasks[[1]], mean)[-1],
tapply(pancreasImages[[2]][,,5], pancreasMasks[[2]], mean)[-1],
tapply(pancreasImages[[3]][,,5], pancreasMasks[[3]], mean)[-1])
expect_equal(assayNames(sce), "counts")
# Intensities
expect_equal(as.numeric(cur_H3), counts(sce)["H3",])
expect_equal(as.numeric(cur_CD99), counts(sce)["CD99",])
expect_equal(as.numeric(cur_PIN), counts(sce)["PIN",])
expect_equal(as.numeric(cur_CD8a), counts(sce)["CD8a",])
expect_equal(as.numeric(cur_CDH), counts(sce)["CDH",])
# Other entries
expect_equal(sce$ImageNb, pancreasSCE$ImageNb)
expect_equal(sce$object_id, pancreasSCE$CellNb)
expect_equal(rownames(sce), rownames(pancreasSCE))
expect_equal(rownames(sce), channelNames(pancreasImages))
# colData entries
expect_equal(names(colData(sce)),
c("ImageNb", "object_id", "s.area", "s.radius.mean",
"m.cx", "m.cy", "m.majoraxis", "m.eccentricity", "ImageName"))
area <- c(table(pancreasMasks[[1]])[-1],
table(pancreasMasks[[2]])[-1],
table(pancreasMasks[[3]])[-1])
expect_equal(sce$s.area, as.numeric(area))
expect_equal(sce$s.radius.mean[20:30],
c(4.636644, 4.764700, 3.166851, 3.637420, 4.538360, 4.423855, 5.031715, 4.556145, 4.854957, 2.911706, 3.895167),
tolerance = 0.0001)
expect_equal(sce$m.cx[20:30],
c(3.776316, 29.804878, 97.243902, 58.478261, 80.092105, 39.873239, 70.260870, 12.092105, 84.920000, 2.588235, 21.271186),
tolerance = 0.0001)
expect_equal(sce$m.cy[20:30],
c(13.71053, 17.71951, 16.12195, 18.56522, 18.36842, 21.67606, 23.11957, 23.78947, 25.02667, 24.50000, 27.27119),
tolerance = 0.0001)
expect_equal(sce$m.majoraxis[20:30],
c(14.651483, 12.792813, 8.718651, 10.983645, 11.665788, 11.333186, 13.358161, 13.151296, 16.216661, 9.080632, 9.137610),
tolerance = 0.0001)
expect_equal(sce$m.eccentricity[20:30],
c(0.8771370, 0.7160534, 0.7079446, 0.8471241, 0.6798998, 0.6019294, 0.7431450, 0.8167983, 0.9096242, 0.8334293, 0.4357446),
tolerance = 0.0001)
# Other standard input
sce <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb",
feature_types = "basic")
expect_equal(names(colData(sce)),
c("ImageNb", "object_id", "ImageName"))
sce <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb",
feature_types = c("basic", "shape"))
expect_equal(names(colData(sce)),
c("ImageNb", "object_id", "s.area", "s.radius.mean", "ImageName"))
sce <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb",
feature_types = c("basic", "shape", "moment"))
expect_equal(names(colData(sce)),
c("ImageNb", "object_id", "s.area", "s.radius.mean",
"m.cx", "m.cy", "m.majoraxis", "m.eccentricity", "ImageName"))
sce <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb",
feature_types = c("basic", "shape", "moment", "haralick"),
haralick_feature = "ent.s1")
expect_equal(names(colData(sce)),
c("ImageNb", "object_id", "s.area", "s.radius.mean",
"m.cx", "m.cy", "m.majoraxis", "m.eccentricity",
"H3.h.ent.s1", "CD99.h.ent.s1", "PIN.h.ent.s1", "CD8a.h.ent.s1", "CDH.h.ent.s1",
"ImageName"))
expect_equal(sce$H3.h.ent.s1[20:30],
c(0.92239897, 0.18210610, 0.00000000, 0.28466019, 0.24327757, 0.28306235, 0.31983314, 0.26878196, 0.05829775, 0.00000000, 0.54267473),
tolerance = 0.0001)
expect_equal(sce$CD99.h.ent.s1[20:30],
c(1.0087171, 0.6879652, 0.0000000, 0.8714867, 0.8615940, 0.6785749, 0.8175975, 0.9346840, 0.7556055, 0.8351088, 0.7464104),
tolerance = 0.0001)
expect_equal(sce$PIN.h.ent.s1[20:30],
c(0.3824506, 0.4245154, 0.7619950, 0.5764059, 0.6587099, 0.5013036, 0.4560755, 0.1652518, 0.7064710, 0.2808083, 0.6687518),
tolerance = 0.0001)
expect_equal(sce$CD8a.h.ent.s1[20:30],
c(0.4807238, 0.3278883, 0.4608923, 0.5163900, 0.4407971, 0.4067465, 0.4099984, 0.5272945, 0.5660128, 0.6652107, 0.4108133),
tolerance = 0.0001)
expect_equal(sce$CDH.h.ent.s1[20:30],
c(0.2464482, 0.9223040, 0.3924199, 0.6494582, 0.7329984, 0.9315071, 0.5924960, 0.7332656, 0.6537887, 0.6607878, 0.7881618),
tolerance = 0.0001)
sce <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb",
feature_types = c("basic", "moment"))
expect_equal(names(colData(sce)),
c("ImageNb", "object_id",
"m.cx", "m.cy", "m.majoraxis", "m.eccentricity", "ImageName"))
sce <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb",
feature_types = c("moment", "basic"))
expect_equal(names(colData(sce)),
c("ImageNb", "object_id",
"m.cx", "m.cy", "m.majoraxis", "m.eccentricity", "ImageName"))
sce <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb",
feature_types = c("basic", "haralick"),
haralick_feature = "ent.s1")
expect_equal(names(colData(sce)),
c("ImageNb", "object_id",
"H3.h.ent.s1", "CD99.h.ent.s1", "PIN.h.ent.s1", "CD8a.h.ent.s1", "CDH.h.ent.s1", "ImageName"))
# Parallelisable
expect_silent(sce.1 <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb",
BPPARAM = BiocParallel::bpparam()))
expect_silent(sce.2 <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb",
BPPARAM = BiocParallel::SerialParam()))
expect_equal(sce.1, sce.2)
expect_equal(as.numeric(cur_H3), counts(sce.1)["H3",])
# different mcols entries work
cur_images <- pancreasImages
mcols(cur_images)$other_entry <- "test"
sce <- measureObjects(pancreasMasks, cur_images, img_id = "ImageNb")
expect_equal(names(colData(sce)),
c("ImageNb", "object_id", "s.area", "s.radius.mean",
"m.cx", "m.cy", "m.majoraxis", "m.eccentricity", "ImageName",
"other_entry"))
cur_masks <- pancreasMasks
mcols(cur_masks)$other_entry_2 <- c("test1", "test2", "test3")
sce <- measureObjects(cur_masks, cur_images, img_id = "ImageNb")
expect_equal(names(colData(sce)),
c("ImageNb", "object_id", "s.area", "s.radius.mean",
"m.cx", "m.cy", "m.majoraxis", "m.eccentricity", "ImageName",
"other_entry", "other_entry_2"))
# Error
expect_error(sce <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb",
feature_types = c("basic", "shape", "moment", "haralick")),
regexp = "Specify at least one haralick feature.",
fixed = TRUE)
expect_error(sce <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb",
feature_types = c("test", "moment")),
regexp = "Only features of type 'basic', 'shape', 'moment' and 'haralick' are allowed.",
fixed = TRUE)
expect_error(sce <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb",
feature_types = c("shape", "moment")),
regexp = "Please specify a basic feature to characterise the marker expression per cell.",
fixed = TRUE)
expect_error(sce <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb",
feature_types = c("basic"), basic_feature = NULL),
regexp = "Please specify a basic feature to characterise the marker expression per cell.",
fixed = TRUE)
expect_error(sce <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb",
feature_types = c("basic"), basic_feature = c("mean", "sd")),
regexp = "Only one intensity feature can be used to characterise the expression of each marker in each object.",
fixed = TRUE)
expect_error(sce <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb",
feature_types = c("basic"), basic_quantiles = "test"),
regexp = "Only numeric quantiles between 0 and 1 allowed.",
fixed = TRUE)
expect_error(sce <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb",
feature_types = c("basic"), basic_feature = "test"),
regexp = "Only basic features of type 'mean', 'sd', 'mad' or the selected quantiles allowed.",
fixed = TRUE)
expect_error(sce <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb",
feature_types = c("basic"), basic_quantiles = 0.2, basic_feature = "q01"),
regexp = "Only basic features of type 'mean', 'sd', 'mad' or the selected quantiles allowed.",
fixed = TRUE)
expect_error(sce <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb",
feature_types = c("basic", "shape"), shape_feature = NULL),
regexp = "Specify at least one shape feature.",
fixed = TRUE)
expect_error(sce <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb",
feature_types = c("basic", "shape"), shape_feature = "test"),
regexp = "Only shape features of type 'area', 'perimeter', 'radius.mean', 'radius.sd', 'radius.max', 'radius.min' allowed.",
fixed = TRUE)
expect_error(sce <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb",
feature_types = c("basic", "shape"), shape_feature = c("area", "test")),
regexp = "Only shape features of type 'area', 'perimeter', 'radius.mean', 'radius.sd', 'radius.max', 'radius.min' allowed.",
fixed = TRUE)
expect_error(sce <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb",
feature_types = c("basic", "moment"), moment_feature = NULL),
regexp = "Specify at least one moment feature.",
fixed = TRUE)
expect_error(sce <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb",
feature_types = c("basic", "moment"), moment_feature = "test"),
regexp = "Only moment features of type 'cx', 'cy', 'majoraxis', 'eccentricity', 'theta' allowed.",
fixed = TRUE)
expect_error(sce <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb",
feature_types = c("basic", "moment"), moment_feature = c("cx", "test")),
regexp = "Only moment features of type 'cx', 'cy', 'majoraxis', 'eccentricity', 'theta' allowed.",
fixed = TRUE)
expect_error(sce <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb",
feature_types = c("basic", "haralick"), haralick_feature = NULL),
regexp = "Specify at least one haralick feature.",
fixed = TRUE)
expect_error(sce <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb",
feature_types = c("basic", "haralick"), haralick_feature = "ent.s1",
haralick_nbins = "test"),
regexp = "Specify the number of bins into which intensity levels are binned.",
fixed = TRUE)
expect_error(sce <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb",
feature_types = c("basic", "haralick"), haralick_feature = "ent.s1",
haralick_nbins = NULL),
regexp = "Specify the number of bins into which intensity levels are binned.",
fixed = TRUE)
expect_error(sce <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb",
feature_types = c("basic", "haralick"), haralick_feature = "ent.s1",
haralick_nbins = c(1, 2)),
regexp = "Specify the number of bins into which intensity levels are binned.",
fixed = TRUE)
expect_error(sce <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb",
feature_types = c("basic", "haralick"), haralick_feature = "ent.s1",
haralick_scales = NULL),
regexp = "Specify the scale (in pixels) over which the haralick features are computed.",
fixed = TRUE)
expect_error(sce <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb",
feature_types = c("basic", "haralick"), haralick_feature = "ent.s1",
haralick_scales = "test"),
regexp = "Specify the scale (in pixels) over which the haralick features are computed.",
fixed = TRUE)
expect_error(sce <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb",
feature_types = c("basic", "haralick"), haralick_feature = "test"),
regexp = "Only haralick features of type asm.s1, con.s1, cor.s1, var.s1, idm.s1, sav.s1, sva.s1, sen.s1, ent.s1, dva.s1, den.s1, f12.s1, f13.s1, asm.s2, con.s2, cor.s2, var.s2, idm.s2, sav.s2, sva.s2, sen.s2, ent.s2, dva.s2, den.s2, f12.s2, f13.s2 allowed.",
fixed = TRUE)
expect_error(sce <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb",
feature_types = c("basic", "haralick"), haralick_feature = c("var.s1", "test")),
regexp = "Only haralick features of type asm.s1, con.s1, cor.s1, var.s1, idm.s1, sav.s1, sva.s1, sen.s1, ent.s1, dva.s1, den.s1, f12.s1, f13.s1, asm.s2, con.s2, cor.s2, var.s2, idm.s2, sav.s2, sva.s2, sen.s2, ent.s2, dva.s2, den.s2, f12.s2, f13.s2 allowed.",
fixed = TRUE)
expect_error(sce <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb",
feature_types = c("basic", "haralick"), haralick_feature = c("var.s1", "test"), haralick_scales = c(1, 3)),
regexp = "Only haralick features of type asm.s1, con.s1, cor.s1, var.s1, idm.s1, sav.s1, sva.s1, sen.s1, ent.s1, dva.s1, den.s1, f12.s1, f13.s1, asm.s3, con.s3, cor.s3, var.s3, idm.s3, sav.s3, sva.s3, sen.s3, ent.s3, dva.s3, den.s3, f12.s3, f13.s3 allowed.",
fixed = TRUE)
# Test for character img_id
expect_silent(sce <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageName"))
expect_true(is.numeric(sce$object_id))
expect_equal(sce$object_id[1:10], c(824, 835, 839, 844, 847, 853, 859, 864, 865, 872))
})
test_that("measureObjects: different settings work", {
data("pancreasSCE")
data("pancreasMasks")
data("pancreasImages")
# Basic features
expect_silent(sce <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb", basic_feature = "sd"))
cur_H3 <- c(tapply(pancreasImages[[1]][,,1], pancreasMasks[[1]], sd)[-1],
tapply(pancreasImages[[2]][,,1], pancreasMasks[[2]], sd)[-1],
tapply(pancreasImages[[3]][,,1], pancreasMasks[[3]], sd)[-1])
cur_H3[is.na(cur_H3)] <- 0
cur_CD99 <- c(tapply(pancreasImages[[1]][,,2], pancreasMasks[[1]], sd)[-1],
tapply(pancreasImages[[2]][,,2], pancreasMasks[[2]], sd)[-1],
tapply(pancreasImages[[3]][,,2], pancreasMasks[[3]], sd)[-1])
cur_CD99[is.na(cur_CD99)] <- 0
cur_PIN <- c(tapply(pancreasImages[[1]][,,3], pancreasMasks[[1]], sd)[-1],
tapply(pancreasImages[[2]][,,3], pancreasMasks[[2]], sd)[-1],
tapply(pancreasImages[[3]][,,3], pancreasMasks[[3]], sd)[-1])
cur_PIN[is.na(cur_PIN)] <- 0
cur_CD8a <- c(tapply(pancreasImages[[1]][,,4], pancreasMasks[[1]], sd)[-1],
tapply(pancreasImages[[2]][,,4], pancreasMasks[[2]], sd)[-1],
tapply(pancreasImages[[3]][,,4], pancreasMasks[[3]], sd)[-1])
cur_CD8a[is.na(cur_CD8a)] <- 0
cur_CDH <- c(tapply(pancreasImages[[1]][,,5], pancreasMasks[[1]], sd)[-1],
tapply(pancreasImages[[2]][,,5], pancreasMasks[[2]], sd)[-1],
tapply(pancreasImages[[3]][,,5], pancreasMasks[[3]], sd)[-1])
cur_CDH[is.na(cur_CDH)] <- 0
expect_equal(assayNames(sce), "counts_sd")
# SD
expect_equal(as.numeric(cur_H3), assay(sce, "counts_sd")["H3",])
expect_equal(as.numeric(cur_CD99), assay(sce, "counts_sd")["CD99",])
expect_equal(as.numeric(cur_PIN), assay(sce, "counts_sd")["PIN",])
expect_equal(as.numeric(cur_CD8a), assay(sce, "counts_sd")["CD8a",])
expect_equal(as.numeric(cur_CDH), assay(sce, "counts_sd")["CDH",])
expect_silent(sce <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb", basic_feature = "mad"))
cur_H3 <- c(tapply(pancreasImages[[1]][,,1], pancreasMasks[[1]], mad)[-1],
tapply(pancreasImages[[2]][,,1], pancreasMasks[[2]], mad)[-1],
tapply(pancreasImages[[3]][,,1], pancreasMasks[[3]], mad)[-1])
cur_H3[is.na(cur_H3)] <- 0
cur_CD99 <- c(tapply(pancreasImages[[1]][,,2], pancreasMasks[[1]], mad)[-1],
tapply(pancreasImages[[2]][,,2], pancreasMasks[[2]], mad)[-1],
tapply(pancreasImages[[3]][,,2], pancreasMasks[[3]], mad)[-1])
cur_CD99[is.na(cur_CD99)] <- 0
cur_PIN <- c(tapply(pancreasImages[[1]][,,3], pancreasMasks[[1]], mad)[-1],
tapply(pancreasImages[[2]][,,3], pancreasMasks[[2]], mad)[-1],
tapply(pancreasImages[[3]][,,3], pancreasMasks[[3]], mad)[-1])
cur_PIN[is.na(cur_PIN)] <- 0
cur_CD8a <- c(tapply(pancreasImages[[1]][,,4], pancreasMasks[[1]], mad)[-1],
tapply(pancreasImages[[2]][,,4], pancreasMasks[[2]], mad)[-1],
tapply(pancreasImages[[3]][,,4], pancreasMasks[[3]], mad)[-1])
cur_CD8a[is.na(cur_CD8a)] <- 0
cur_CDH <- c(tapply(pancreasImages[[1]][,,5], pancreasMasks[[1]], mad)[-1],
tapply(pancreasImages[[2]][,,5], pancreasMasks[[2]], mad)[-1],
tapply(pancreasImages[[3]][,,5], pancreasMasks[[3]], mad)[-1])
cur_CDH[is.na(cur_CDH)] <- 0
expect_equal(assayNames(sce), "counts_mad")
# MAD
expect_equal(as.numeric(cur_H3), assay(sce, "counts_mad")["H3",])
expect_equal(as.numeric(cur_CD99), assay(sce, "counts_mad")["CD99",])
expect_equal(as.numeric(cur_PIN), assay(sce, "counts_mad")["PIN",])
expect_equal(as.numeric(cur_CD8a), assay(sce, "counts_mad")["CD8a",])
expect_equal(as.numeric(cur_CDH), assay(sce, "counts_mad")["CDH",])
expect_silent(sce <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb", basic_feature = "q05", basic_quantiles = 0.5))
cur_H3 <- c(tapply(pancreasImages[[1]][,,1], pancreasMasks[[1]], median)[-1],
tapply(pancreasImages[[2]][,,1], pancreasMasks[[2]], median)[-1],
tapply(pancreasImages[[3]][,,1], pancreasMasks[[3]], median)[-1])
cur_CD99 <- c(tapply(pancreasImages[[1]][,,2], pancreasMasks[[1]], median)[-1],
tapply(pancreasImages[[2]][,,2], pancreasMasks[[2]], median)[-1],
tapply(pancreasImages[[3]][,,2], pancreasMasks[[3]], median)[-1])
cur_PIN <- c(tapply(pancreasImages[[1]][,,3], pancreasMasks[[1]], median)[-1],
tapply(pancreasImages[[2]][,,3], pancreasMasks[[2]], median)[-1],
tapply(pancreasImages[[3]][,,3], pancreasMasks[[3]], median)[-1])
cur_CD8a <- c(tapply(pancreasImages[[1]][,,4], pancreasMasks[[1]], median)[-1],
tapply(pancreasImages[[2]][,,4], pancreasMasks[[2]], median)[-1],
tapply(pancreasImages[[3]][,,4], pancreasMasks[[3]], median)[-1])
cur_CDH <- c(tapply(pancreasImages[[1]][,,5], pancreasMasks[[1]], median)[-1],
tapply(pancreasImages[[2]][,,5], pancreasMasks[[2]], median)[-1],
tapply(pancreasImages[[3]][,,5], pancreasMasks[[3]], median)[-1])
expect_equal(assayNames(sce), "counts_q05")
# Median
expect_equal(as.numeric(cur_H3), assay(sce, "counts_q05")["H3",])
expect_equal(as.numeric(cur_CD99), assay(sce, "counts_q05")["CD99",])
expect_equal(as.numeric(cur_PIN), assay(sce, "counts_q05")["PIN",])
expect_equal(as.numeric(cur_CD8a), assay(sce, "counts_q05")["CD8a",])
expect_equal(as.numeric(cur_CDH), assay(sce, "counts_q05")["CDH",])
# Shape features
expect_silent(sce <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb", feature_types = c("basic", "shape"),
shape_feature = c('area', 'perimeter', 'radius.mean', 'radius.sd', 'radius.max', 'radius.min')))
expect_equal(names(colData(sce)), c("ImageNb", "object_id", "s.area", "s.perimeter", "s.radius.mean", "s.radius.sd", "s.radius.min", "s.radius.max", "ImageName"))
expect_equal(sce$s.perimeter[40:50], c(28, 39, 25, 10, 16, 15, 13, 22, 31, 37, 5))
expect_equal(sce$s.radius.sd[40:50], c(0.6381883, 1.2634757, 0.5599216, 0.5707716, 0.6687795, 0.4334581, 0.7503223, 0.4679423, 0.7259625, 0.8618819, 0.2850134), tolerance = 0.00001)
expect_equal(sce$s.radius.min[40:50], c(3.5959870, 3.1786599, 2.9772471, 0.7810250, 1.4710753, 1.2292726, 0.8770580, 2.8477191, 3.2325735, 3.9868300, 0.4472136), tolerance = 0.00001)
expect_equal(sce$s.radius.max[40:50], c(6.123933, 8.641064, 5.209990, 2.570992, 3.628231, 2.996294, 3.328201, 4.788286, 6.002514, 7.190662, 1.264911), tolerance = 0.00001)
# Moment features
expect_silent(sce <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb", feature_types = c("basic", "moment"),
moment_feature = c('cx', 'cy', 'majoraxis', 'eccentricity', 'theta')))
expect_equal(names(colData(sce)), c("ImageNb", "object_id", "m.cx", "m.cy", "m.majoraxis", "m.eccentricity", "m.theta", "ImageName"))
expect_equal(sce$m.theta[40:50], c(0.60960841, -0.04246836, -0.68045282, 1.57079633, 0.89866183, -1.11289559, -1.10737273, -0.96767979, 1.53796956, -0.09403954, -1.24904577))
# Haralick features
expect_silent(sce <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb", feature_types = c("basic", "haralick"),
haralick_feature = c("asm.s1", "con.s1", "cor.s1", "var.s1", "idm.s1", "sav.s1", "sva.s1", "sen.s1", "ent.s1", "dva.s1", "den.s1", "f12.s1", "f13.s1", "asm.s2", "con.s2", "cor.s2", "var.s2", "idm.s2", "sav.s2", "sva.s2", "sen.s2", "ent.s2", "dva.s2", "den.s2", "f12.s2", "f13.s2")))
features <- c("asm.s1", "con.s1", "cor.s1", "var.s1", "idm.s1", "sav.s1", "sva.s1", "sen.s1", "ent.s1", "dva.s1", "den.s1", "f12.s1", "f13.s1", "asm.s2", "con.s2", "cor.s2", "var.s2", "idm.s2", "sav.s2", "sva.s2", "sen.s2", "ent.s2", "dva.s2", "den.s2", "f12.s2", "f13.s2")
features <- paste0("h.", features)
features <- paste0(rep(channelNames(pancreasImages), each = length(features)), ".", features)
features <- c("ImageNb", "object_id", features, "ImageName")
expect_equal(names(colData(sce)), features)
cur_basic <- computeFeatures.basic(pancreasMasks[[1]], pancreasImages[[1]][,,1])
cur_haralick <- computeFeatures.haralick(pancreasMasks[[1]], pancreasImages[[1]][,,1])
cur_haralick <- cur_haralick[as.numeric(rownames(cur_basic)),]
colnames(cur_haralick) <- paste0("H3.", colnames(cur_haralick))
expect_equal(as.matrix(colData(sce)[sce$ImageNb == 1,colnames(cur_haralick)]), cur_haralick)
cur_haralick <- computeFeatures.haralick(pancreasMasks[[1]], pancreasImages[[1]][,,2])
cur_haralick <- cur_haralick[as.numeric(rownames(cur_basic)),]
colnames(cur_haralick) <- paste0("CD99.", colnames(cur_haralick))
expect_equal(as.matrix(colData(sce)[sce$ImageNb == 1,colnames(cur_haralick)]), cur_haralick)
expect_silent(sce <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb", feature_types = c("basic", "haralick"), haralick_scales = c(1,4),
haralick_feature = c("asm.s1", "con.s1", "cor.s1", "var.s1", "idm.s1", "sav.s1", "sva.s1", "sen.s1", "ent.s1", "dva.s1", "den.s1", "f12.s1", "f13.s1", "asm.s4", "con.s4", "cor.s4", "var.s4", "idm.s4", "sav.s4", "sva.s4", "sen.s4", "ent.s4", "dva.s4", "den.s4", "f12.s4", "f13.s4")))
features <- c("asm.s1", "con.s1", "cor.s1", "var.s1", "idm.s1", "sav.s1", "sva.s1", "sen.s1", "ent.s1", "dva.s1", "den.s1", "f12.s1", "f13.s1", "asm.s4", "con.s4", "cor.s4", "var.s4", "idm.s4", "sav.s4", "sva.s4", "sen.s4", "ent.s4", "dva.s4", "den.s4", "f12.s4", "f13.s4")
features <- paste0("h.", features)
features <- paste0(rep(channelNames(pancreasImages), each = length(features)), ".", features)
features <- c("ImageNb", "object_id", features, "ImageName")
expect_equal(names(colData(sce)), features)
cur_basic <- computeFeatures.basic(pancreasMasks[[1]], pancreasImages[[1]][,,1])
cur_haralick <- computeFeatures.haralick(pancreasMasks[[1]], pancreasImages[[1]][,,1], haralick.scales = c(1,4))
cur_haralick <- cur_haralick[as.numeric(rownames(cur_basic)),]
colnames(cur_haralick) <- paste0("H3.", colnames(cur_haralick))
expect_equal(as.matrix(colData(sce)[sce$ImageNb == 1,colnames(cur_haralick)]), cur_haralick)
cur_haralick <- computeFeatures.haralick(pancreasMasks[[1]], pancreasImages[[1]][,,2], haralick.scales = c(1,4))
cur_haralick <- cur_haralick[as.numeric(rownames(cur_basic)),]
colnames(cur_haralick) <- paste0("CD99.", colnames(cur_haralick))
expect_equal(as.matrix(colData(sce)[sce$ImageNb == 1,colnames(cur_haralick)]), cur_haralick)
expect_silent(sce <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb", feature_types = c("basic", "haralick"), haralick_nbins = 16,
haralick_feature = c("asm.s1", "con.s1", "cor.s1", "var.s1", "idm.s1", "sav.s1", "sva.s1", "sen.s1", "ent.s1", "dva.s1", "den.s1", "f12.s1", "f13.s1", "asm.s2", "con.s2", "cor.s2", "var.s2", "idm.s2", "sav.s2", "sva.s2", "sen.s2", "ent.s2", "dva.s2", "den.s2", "f12.s2", "f13.s2")))
features <- c("asm.s1", "con.s1", "cor.s1", "var.s1", "idm.s1", "sav.s1", "sva.s1", "sen.s1", "ent.s1", "dva.s1", "den.s1", "f12.s1", "f13.s1", "asm.s2", "con.s2", "cor.s2", "var.s2", "idm.s2", "sav.s2", "sva.s2", "sen.s2", "ent.s2", "dva.s2", "den.s2", "f12.s2", "f13.s2")
features <- paste0("h.", features)
features <- paste0(rep(channelNames(pancreasImages), each = length(features)), ".", features)
features <- c("ImageNb", "object_id", features, "ImageName")
expect_equal(names(colData(sce)), features)
cur_basic <- computeFeatures.basic(pancreasMasks[[1]], pancreasImages[[1]][,,1])
cur_haralick <- computeFeatures.haralick(pancreasMasks[[1]], pancreasImages[[1]][,,1], haralick.nbins = 16)
cur_haralick <- cur_haralick[as.numeric(rownames(cur_basic)),]
colnames(cur_haralick) <- paste0("H3.", colnames(cur_haralick))
expect_equal(as.matrix(colData(sce)[sce$ImageNb == 1,colnames(cur_haralick)]), cur_haralick)
cur_haralick <- computeFeatures.haralick(pancreasMasks[[1]], pancreasImages[[1]][,,2], haralick.nbins = 16)
cur_haralick <- cur_haralick[as.numeric(rownames(cur_basic)),]
colnames(cur_haralick) <- paste0("CD99.", colnames(cur_haralick))
expect_equal(as.matrix(colData(sce)[sce$ImageNb == 1,colnames(cur_haralick)]), cur_haralick)
expect_silent(sce <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb", feature_types = c("basic", "haralick"),
haralick_feature = c("asm.s1", "ent.s2")))
features <- c("asm.s1", "ent.s2")
features <- paste0("h.", features)
features <- paste0(rep(channelNames(pancreasImages), each = length(features)), ".", features)
features <- c("ImageNb", "object_id", features, "ImageName")
expect_equal(names(colData(sce)), features)
cur_basic <- computeFeatures.basic(pancreasMasks[[1]], pancreasImages[[1]][,,1])
cur_haralick <- computeFeatures.haralick(pancreasMasks[[1]], pancreasImages[[1]][,,1])
cur_haralick <- cur_haralick[as.numeric(rownames(cur_basic)),]
colnames(cur_haralick) <- paste0("H3.", colnames(cur_haralick))
expect_equal(as.matrix(colData(sce)[sce$ImageNb == 1,c("H3.h.asm.s1", "H3.h.ent.s2")]), cur_haralick[,c("H3.h.asm.s1", "H3.h.ent.s2")])
cur_haralick <- computeFeatures.haralick(pancreasMasks[[1]], pancreasImages[[1]][,,2])
cur_haralick <- cur_haralick[as.numeric(rownames(cur_basic)),]
colnames(cur_haralick) <- paste0("CD99.", colnames(cur_haralick))
expect_equal(as.matrix(colData(sce)[sce$ImageNb == 1,c("CD99.h.asm.s1", "CD99.h.ent.s2")]), cur_haralick[,c("CD99.h.asm.s1", "CD99.h.ent.s2")])
})
test_that("measureObjects: defaults work", {
data("pancreasSCE")
data("pancreasMasks")
data("pancreasImages")
# Works
expect_silent(spe <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb", return_as = "spe"))
cur_H3 <- c(tapply(pancreasImages[[1]][,,1], pancreasMasks[[1]], mean)[-1],
tapply(pancreasImages[[2]][,,1], pancreasMasks[[2]], mean)[-1],
tapply(pancreasImages[[3]][,,1], pancreasMasks[[3]], mean)[-1])
cur_CD99 <- c(tapply(pancreasImages[[1]][,,2], pancreasMasks[[1]], mean)[-1],
tapply(pancreasImages[[2]][,,2], pancreasMasks[[2]], mean)[-1],
tapply(pancreasImages[[3]][,,2], pancreasMasks[[3]], mean)[-1])
cur_PIN <- c(tapply(pancreasImages[[1]][,,3], pancreasMasks[[1]], mean)[-1],
tapply(pancreasImages[[2]][,,3], pancreasMasks[[2]], mean)[-1],
tapply(pancreasImages[[3]][,,3], pancreasMasks[[3]], mean)[-1])
cur_CD8a <- c(tapply(pancreasImages[[1]][,,4], pancreasMasks[[1]], mean)[-1],
tapply(pancreasImages[[2]][,,4], pancreasMasks[[2]], mean)[-1],
tapply(pancreasImages[[3]][,,4], pancreasMasks[[3]], mean)[-1])
cur_CDH <- c(tapply(pancreasImages[[1]][,,5], pancreasMasks[[1]], mean)[-1],
tapply(pancreasImages[[2]][,,5], pancreasMasks[[2]], mean)[-1],
tapply(pancreasImages[[3]][,,5], pancreasMasks[[3]], mean)[-1])
expect_equal(assayNames(spe), "counts")
# Intensities
expect_equal(as.numeric(cur_H3), counts(spe)["H3",])
expect_equal(as.numeric(cur_CD99), counts(spe)["CD99",])
expect_equal(as.numeric(cur_PIN), counts(spe)["PIN",])
expect_equal(as.numeric(cur_CD8a), counts(spe)["CD8a",])
expect_equal(as.numeric(cur_CDH), counts(spe)["CDH",])
# Other entries
expect_equal(spe$ImageNb, pancreasSCE$ImageNb)
expect_equal(spe$object_id, pancreasSCE$CellNb)
expect_equal(rownames(spe), rownames(pancreasSCE))
expect_equal(rownames(spe), channelNames(pancreasImages))
# colData entries
expect_equal(names(colData(spe)),
c("ImageNb", "object_id", "s.area", "s.radius.mean",
"m.majoraxis", "m.eccentricity", "sample_id", "ImageName"))
area <- c(table(pancreasMasks[[1]])[-1],
table(pancreasMasks[[2]])[-1],
table(pancreasMasks[[3]])[-1])
expect_equal(spe$s.area, as.numeric(area))
expect_equal(spe$s.radius.mean[20:30],
c(4.636644, 4.764700, 3.166851, 3.637420, 4.538360, 4.423855, 5.031715, 4.556145, 4.854957, 2.911706, 3.895167),
tolerance = 0.0001)
expect_equal(SpatialExperiment::spatialCoords(spe)[20:30,"m.cx"],
c(3.776316, 29.804878, 97.243902, 58.478261, 80.092105, 39.873239, 70.260870, 12.092105, 84.920000, 2.588235, 21.271186),
tolerance = 0.0001)
expect_equal(SpatialExperiment::spatialCoords(spe)[20:30,"m.cy"],
c(13.71053, 17.71951, 16.12195, 18.56522, 18.36842, 21.67606, 23.11957, 23.78947, 25.02667, 24.50000, 27.27119),
tolerance = 0.0001)
expect_equal(spe$m.majoraxis[20:30],
c(14.651483, 12.792813, 8.718651, 10.983645, 11.665788, 11.333186, 13.358161, 13.151296, 16.216661, 9.080632, 9.137610),
tolerance = 0.0001)
expect_equal(spe$m.eccentricity[20:30],
c(0.8771370, 0.7160534, 0.7079446, 0.8471241, 0.6798998, 0.6019294, 0.7431450, 0.8167983, 0.9096242, 0.8334293, 0.4357446),
tolerance = 0.0001)
# Other standard input
spe <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb",
feature_types = "basic", return_as = "spe")
expect_equal(names(colData(spe)),
c("ImageNb", "object_id", "sample_id", "ImageName"))
spe <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb",
feature_types = c("basic", "shape"), return_as = "spe")
expect_equal(names(colData(spe)),
c("ImageNb", "object_id", "s.area", "s.radius.mean", "sample_id", "ImageName"))
spe <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb",
feature_types = c("basic", "shape", "moment"), return_as = "spe")
expect_equal(names(colData(spe)),
c("ImageNb", "object_id", "s.area", "s.radius.mean",
"m.majoraxis", "m.eccentricity", "sample_id", "ImageName"))
expect_equal(colnames(SpatialExperiment::spatialCoords(spe)),
c("m.cx", "m.cy"))
spe <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb",
feature_types = c("basic", "shape", "moment", "haralick"),
haralick_feature = "ent.s1", return_as = "spe")
expect_equal(names(colData(spe)),
c("ImageNb", "object_id", "s.area", "s.radius.mean",
"m.majoraxis", "m.eccentricity",
"H3.h.ent.s1", "CD99.h.ent.s1", "PIN.h.ent.s1", "CD8a.h.ent.s1", "CDH.h.ent.s1",
"sample_id", "ImageName"))
expect_equal(spe$H3.h.ent.s1[20:30],
c(0.92239897, 0.18210610, 0.00000000, 0.28466019, 0.24327757, 0.28306235, 0.31983314, 0.26878196, 0.05829775, 0.00000000, 0.54267473),
tolerance = 0.0001)
expect_equal(spe$CD99.h.ent.s1[20:30],
c(1.0087171, 0.6879652, 0.0000000, 0.8714867, 0.8615940, 0.6785749, 0.8175975, 0.9346840, 0.7556055, 0.8351088, 0.7464104),
tolerance = 0.0001)
expect_equal(spe$PIN.h.ent.s1[20:30],
c(0.3824506, 0.4245154, 0.7619950, 0.5764059, 0.6587099, 0.5013036, 0.4560755, 0.1652518, 0.7064710, 0.2808083, 0.6687518),
tolerance = 0.0001)
expect_equal(spe$CD8a.h.ent.s1[20:30],
c(0.4807238, 0.3278883, 0.4608923, 0.5163900, 0.4407971, 0.4067465, 0.4099984, 0.5272945, 0.5660128, 0.6652107, 0.4108133),
tolerance = 0.0001)
expect_equal(spe$CDH.h.ent.s1[20:30],
c(0.2464482, 0.9223040, 0.3924199, 0.6494582, 0.7329984, 0.9315071, 0.5924960, 0.7332656, 0.6537887, 0.6607878, 0.7881618),
tolerance = 0.0001)
spe <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb",
feature_types = c("basic", "moment"), return_as = "spe")
expect_equal(names(colData(spe)),
c("ImageNb", "object_id",
"m.majoraxis", "m.eccentricity", "sample_id", "ImageName"))
spe <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb",
feature_types = c("moment", "basic"), return_as = "spe")
expect_equal(names(colData(spe)),
c("ImageNb", "object_id",
"m.majoraxis", "m.eccentricity", "sample_id", "ImageName"))
spe <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb",
feature_types = c("basic", "haralick"),
haralick_feature = "ent.s1", return_as = "spe")
expect_equal(names(colData(spe)),
c("ImageNb", "object_id",
"H3.h.ent.s1", "CD99.h.ent.s1", "PIN.h.ent.s1", "CD8a.h.ent.s1", "CDH.h.ent.s1", "sample_id", "ImageName"))
# Parallelisable
expect_silent(spe.1 <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb", return_as = "spe",
BPPARAM = BiocParallel::bpparam()))
expect_silent(spe.2 <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb", return_as = "spe",
BPPARAM = BiocParallel::SerialParam()))
expect_equal(spe.1, spe.2)
expect_equal(as.numeric(cur_H3), counts(spe.1)["H3",])
# different mcols entries work
cur_images <- pancreasImages
mcols(cur_images)$other_entry <- "test"
spe <- measureObjects(pancreasMasks, cur_images, img_id = "ImageNb", return_as = "spe")
expect_equal(names(colData(spe)),
c("ImageNb", "object_id", "s.area", "s.radius.mean",
"m.majoraxis", "m.eccentricity", "sample_id", "ImageName",
"other_entry"))
cur_masks <- pancreasMasks
mcols(cur_masks)$other_entry_2 <- c("test1", "test2", "test3")
spe <- measureObjects(cur_masks, cur_images, img_id = "ImageNb", return_as = "spe")
expect_equal(names(colData(spe)),
c("ImageNb", "object_id", "s.area", "s.radius.mean",
"m.majoraxis", "m.eccentricity", "sample_id", "ImageName",
"other_entry", "other_entry_2"))
})
BodenmillerGroup/SingleCellMapper documentation built on July 2, 2024, 4:31 p.m.
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
test_that("measureObjects: defaults work", {
data("pancreasSCE")
data("pancreasMasks")
data("pancreasImages")
# Works
expect_silent(sce <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb"))
cur_H3 <- c(tapply(pancreasImages[[1]][,,1], pancreasMasks[[1]], mean)[-1],
tapply(pancreasImages[[2]][,,1], pancreasMasks[[2]], mean)[-1],
tapply(pancreasImages[[3]][,,1], pancreasMasks[[3]], mean)[-1])
cur_CD99 <- c(tapply(pancreasImages[[1]][,,2], pancreasMasks[[1]], mean)[-1],
tapply(pancreasImages[[2]][,,2], pancreasMasks[[2]], mean)[-1],
tapply(pancreasImages[[3]][,,2], pancreasMasks[[3]], mean)[-1])
cur_PIN <- c(tapply(pancreasImages[[1]][,,3], pancreasMasks[[1]], mean)[-1],
tapply(pancreasImages[[2]][,,3], pancreasMasks[[2]], mean)[-1],
tapply(pancreasImages[[3]][,,3], pancreasMasks[[3]], mean)[-1])
cur_CD8a <- c(tapply(pancreasImages[[1]][,,4], pancreasMasks[[1]], mean)[-1],
tapply(pancreasImages[[2]][,,4], pancreasMasks[[2]], mean)[-1],
tapply(pancreasImages[[3]][,,4], pancreasMasks[[3]], mean)[-1])
cur_CDH <- c(tapply(pancreasImages[[1]][,,5], pancreasMasks[[1]], mean)[-1],
tapply(pancreasImages[[2]][,,5], pancreasMasks[[2]], mean)[-1],
tapply(pancreasImages[[3]][,,5], pancreasMasks[[3]], mean)[-1])
expect_equal(assayNames(sce), "counts")
# Intensities
expect_equal(as.numeric(cur_H3), counts(sce)["H3",])
expect_equal(as.numeric(cur_CD99), counts(sce)["CD99",])
expect_equal(as.numeric(cur_PIN), counts(sce)["PIN",])
expect_equal(as.numeric(cur_CD8a), counts(sce)["CD8a",])
expect_equal(as.numeric(cur_CDH), counts(sce)["CDH",])
# Other entries
expect_equal(sce$ImageNb, pancreasSCE$ImageNb)
expect_equal(sce$object_id, pancreasSCE$CellNb)
expect_equal(rownames(sce), rownames(pancreasSCE))
expect_equal(rownames(sce), channelNames(pancreasImages))
# colData entries
expect_equal(names(colData(sce)),
c("ImageNb", "object_id", "s.area", "s.radius.mean",
"m.cx", "m.cy", "m.majoraxis", "m.eccentricity", "ImageName"))
area <- c(table(pancreasMasks[[1]])[-1],
table(pancreasMasks[[2]])[-1],
table(pancreasMasks[[3]])[-1])
expect_equal(sce$s.area, as.numeric(area))
expect_equal(sce$s.radius.mean[20:30],
c(4.636644, 4.764700, 3.166851, 3.637420, 4.538360, 4.423855, 5.031715, 4.556145, 4.854957, 2.911706, 3.895167),
tolerance = 0.0001)
expect_equal(sce$m.cx[20:30],
c(3.776316, 29.804878, 97.243902, 58.478261, 80.092105, 39.873239, 70.260870, 12.092105, 84.920000, 2.588235, 21.271186),
tolerance = 0.0001)
expect_equal(sce$m.cy[20:30],
c(13.71053, 17.71951, 16.12195, 18.56522, 18.36842, 21.67606, 23.11957, 23.78947, 25.02667, 24.50000, 27.27119),
tolerance = 0.0001)
expect_equal(sce$m.majoraxis[20:30],
c(14.651483, 12.792813, 8.718651, 10.983645, 11.665788, 11.333186, 13.358161, 13.151296, 16.216661, 9.080632, 9.137610),
tolerance = 0.0001)
expect_equal(sce$m.eccentricity[20:30],
c(0.8771370, 0.7160534, 0.7079446, 0.8471241, 0.6798998, 0.6019294, 0.7431450, 0.8167983, 0.9096242, 0.8334293, 0.4357446),
tolerance = 0.0001)
# Other standard input
sce <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb",
feature_types = "basic")
expect_equal(names(colData(sce)),
c("ImageNb", "object_id", "ImageName"))
sce <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb",
feature_types = c("basic", "shape"))
expect_equal(names(colData(sce)),
c("ImageNb", "object_id", "s.area", "s.radius.mean", "ImageName"))
sce <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb",
feature_types = c("basic", "shape", "moment"))
expect_equal(names(colData(sce)),
c("ImageNb", "object_id", "s.area", "s.radius.mean",
"m.cx", "m.cy", "m.majoraxis", "m.eccentricity", "ImageName"))
sce <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb",
feature_types = c("basic", "shape", "moment", "haralick"),
haralick_feature = "ent.s1")
expect_equal(names(colData(sce)),
c("ImageNb", "object_id", "s.area", "s.radius.mean",
"m.cx", "m.cy", "m.majoraxis", "m.eccentricity",
"H3.h.ent.s1", "CD99.h.ent.s1", "PIN.h.ent.s1", "CD8a.h.ent.s1", "CDH.h.ent.s1",
"ImageName"))
expect_equal(sce$H3.h.ent.s1[20:30],
c(0.92239897, 0.18210610, 0.00000000, 0.28466019, 0.24327757, 0.28306235, 0.31983314, 0.26878196, 0.05829775, 0.00000000, 0.54267473),
tolerance = 0.0001)
expect_equal(sce$CD99.h.ent.s1[20:30],
c(1.0087171, 0.6879652, 0.0000000, 0.8714867, 0.8615940, 0.6785749, 0.8175975, 0.9346840, 0.7556055, 0.8351088, 0.7464104),
tolerance = 0.0001)
expect_equal(sce$PIN.h.ent.s1[20:30],
c(0.3824506, 0.4245154, 0.7619950, 0.5764059, 0.6587099, 0.5013036, 0.4560755, 0.1652518, 0.7064710, 0.2808083, 0.6687518),
tolerance = 0.0001)
expect_equal(sce$CD8a.h.ent.s1[20:30],
c(0.4807238, 0.3278883, 0.4608923, 0.5163900, 0.4407971, 0.4067465, 0.4099984, 0.5272945, 0.5660128, 0.6652107, 0.4108133),
tolerance = 0.0001)
expect_equal(sce$CDH.h.ent.s1[20:30],
c(0.2464482, 0.9223040, 0.3924199, 0.6494582, 0.7329984, 0.9315071, 0.5924960, 0.7332656, 0.6537887, 0.6607878, 0.7881618),
tolerance = 0.0001)
sce <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb",
feature_types = c("basic", "moment"))
expect_equal(names(colData(sce)),
c("ImageNb", "object_id",
"m.cx", "m.cy", "m.majoraxis", "m.eccentricity", "ImageName"))
sce <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb",
feature_types = c("moment", "basic"))
expect_equal(names(colData(sce)),
c("ImageNb", "object_id",
"m.cx", "m.cy", "m.majoraxis", "m.eccentricity", "ImageName"))
sce <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb",
feature_types = c("basic", "haralick"),
haralick_feature = "ent.s1")
expect_equal(names(colData(sce)),
c("ImageNb", "object_id",
"H3.h.ent.s1", "CD99.h.ent.s1", "PIN.h.ent.s1", "CD8a.h.ent.s1", "CDH.h.ent.s1", "ImageName"))
# Parallelisable
expect_silent(sce.1 <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb",
BPPARAM = BiocParallel::bpparam()))
expect_silent(sce.2 <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb",
BPPARAM = BiocParallel::SerialParam()))
expect_equal(sce.1, sce.2)
expect_equal(as.numeric(cur_H3), counts(sce.1)["H3",])
# different mcols entries work
cur_images <- pancreasImages
mcols(cur_images)$other_entry <- "test"
sce <- measureObjects(pancreasMasks, cur_images, img_id = "ImageNb")
expect_equal(names(colData(sce)),
c("ImageNb", "object_id", "s.area", "s.radius.mean",
"m.cx", "m.cy", "m.majoraxis", "m.eccentricity", "ImageName",
"other_entry"))
cur_masks <- pancreasMasks
mcols(cur_masks)$other_entry_2 <- c("test1", "test2", "test3")
sce <- measureObjects(cur_masks, cur_images, img_id = "ImageNb")
expect_equal(names(colData(sce)),
c("ImageNb", "object_id", "s.area", "s.radius.mean",
"m.cx", "m.cy", "m.majoraxis", "m.eccentricity", "ImageName",
"other_entry", "other_entry_2"))
# Error
expect_error(sce <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb",
feature_types = c("basic", "shape", "moment", "haralick")),
regexp = "Specify at least one haralick feature.",
fixed = TRUE)
expect_error(sce <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb",
feature_types = c("test", "moment")),
regexp = "Only features of type 'basic', 'shape', 'moment' and 'haralick' are allowed.",
fixed = TRUE)
expect_error(sce <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb",
feature_types = c("shape", "moment")),
regexp = "Please specify a basic feature to characterise the marker expression per cell.",
fixed = TRUE)
expect_error(sce <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb",
feature_types = c("basic"), basic_feature = NULL),
regexp = "Please specify a basic feature to characterise the marker expression per cell.",
fixed = TRUE)
expect_error(sce <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb",
feature_types = c("basic"), basic_feature = c("mean", "sd")),
regexp = "Only one intensity feature can be used to characterise the expression of each marker in each object.",
fixed = TRUE)
expect_error(sce <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb",
feature_types = c("basic"), basic_quantiles = "test"),
regexp = "Only numeric quantiles between 0 and 1 allowed.",
fixed = TRUE)
expect_error(sce <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb",
feature_types = c("basic"), basic_feature = "test"),
regexp = "Only basic features of type 'mean', 'sd', 'mad' or the selected quantiles allowed.",
fixed = TRUE)
expect_error(sce <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb",
feature_types = c("basic"), basic_quantiles = 0.2, basic_feature = "q01"),
regexp = "Only basic features of type 'mean', 'sd', 'mad' or the selected quantiles allowed.",
fixed = TRUE)
expect_error(sce <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb",
feature_types = c("basic", "shape"), shape_feature = NULL),
regexp = "Specify at least one shape feature.",
fixed = TRUE)
expect_error(sce <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb",
feature_types = c("basic", "shape"), shape_feature = "test"),
regexp = "Only shape features of type 'area', 'perimeter', 'radius.mean', 'radius.sd', 'radius.max', 'radius.min' allowed.",
fixed = TRUE)
expect_error(sce <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb",
feature_types = c("basic", "shape"), shape_feature = c("area", "test")),
regexp = "Only shape features of type 'area', 'perimeter', 'radius.mean', 'radius.sd', 'radius.max', 'radius.min' allowed.",
fixed = TRUE)
expect_error(sce <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb",
feature_types = c("basic", "moment"), moment_feature = NULL),
regexp = "Specify at least one moment feature.",
fixed = TRUE)
expect_error(sce <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb",
feature_types = c("basic", "moment"), moment_feature = "test"),
regexp = "Only moment features of type 'cx', 'cy', 'majoraxis', 'eccentricity', 'theta' allowed.",
fixed = TRUE)
expect_error(sce <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb",
feature_types = c("basic", "moment"), moment_feature = c("cx", "test")),
regexp = "Only moment features of type 'cx', 'cy', 'majoraxis', 'eccentricity', 'theta' allowed.",
fixed = TRUE)
expect_error(sce <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb",
feature_types = c("basic", "haralick"), haralick_feature = NULL),
regexp = "Specify at least one haralick feature.",
fixed = TRUE)
expect_error(sce <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb",
feature_types = c("basic", "haralick"), haralick_feature = "ent.s1",
haralick_nbins = "test"),
regexp = "Specify the number of bins into which intensity levels are binned.",
fixed = TRUE)
expect_error(sce <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb",
feature_types = c("basic", "haralick"), haralick_feature = "ent.s1",
haralick_nbins = NULL),
regexp = "Specify the number of bins into which intensity levels are binned.",
fixed = TRUE)
expect_error(sce <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb",
feature_types = c("basic", "haralick"), haralick_feature = "ent.s1",
haralick_nbins = c(1, 2)),
regexp = "Specify the number of bins into which intensity levels are binned.",
fixed = TRUE)
expect_error(sce <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb",
feature_types = c("basic", "haralick"), haralick_feature = "ent.s1",
haralick_scales = NULL),
regexp = "Specify the scale (in pixels) over which the haralick features are computed.",
fixed = TRUE)
expect_error(sce <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb",
feature_types = c("basic", "haralick"), haralick_feature = "ent.s1",
haralick_scales = "test"),
regexp = "Specify the scale (in pixels) over which the haralick features are computed.",
fixed = TRUE)
expect_error(sce <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb",
feature_types = c("basic", "haralick"), haralick_feature = "test"),
regexp = "Only haralick features of type asm.s1, con.s1, cor.s1, var.s1, idm.s1, sav.s1, sva.s1, sen.s1, ent.s1, dva.s1, den.s1, f12.s1, f13.s1, asm.s2, con.s2, cor.s2, var.s2, idm.s2, sav.s2, sva.s2, sen.s2, ent.s2, dva.s2, den.s2, f12.s2, f13.s2 allowed.",
fixed = TRUE)
expect_error(sce <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb",
feature_types = c("basic", "haralick"), haralick_feature = c("var.s1", "test")),
regexp = "Only haralick features of type asm.s1, con.s1, cor.s1, var.s1, idm.s1, sav.s1, sva.s1, sen.s1, ent.s1, dva.s1, den.s1, f12.s1, f13.s1, asm.s2, con.s2, cor.s2, var.s2, idm.s2, sav.s2, sva.s2, sen.s2, ent.s2, dva.s2, den.s2, f12.s2, f13.s2 allowed.",
fixed = TRUE)
expect_error(sce <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb",
feature_types = c("basic", "haralick"), haralick_feature = c("var.s1", "test"), haralick_scales = c(1, 3)),
regexp = "Only haralick features of type asm.s1, con.s1, cor.s1, var.s1, idm.s1, sav.s1, sva.s1, sen.s1, ent.s1, dva.s1, den.s1, f12.s1, f13.s1, asm.s3, con.s3, cor.s3, var.s3, idm.s3, sav.s3, sva.s3, sen.s3, ent.s3, dva.s3, den.s3, f12.s3, f13.s3 allowed.",
fixed = TRUE)
# Test for character img_id
expect_silent(sce <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageName"))
expect_true(is.numeric(sce$object_id))
expect_equal(sce$object_id[1:10], c(824, 835, 839, 844, 847, 853, 859, 864, 865, 872))
})
test_that("measureObjects: different settings work", {
data("pancreasSCE")
data("pancreasMasks")
data("pancreasImages")
# Basic features
expect_silent(sce <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb", basic_feature = "sd"))
cur_H3 <- c(tapply(pancreasImages[[1]][,,1], pancreasMasks[[1]], sd)[-1],
tapply(pancreasImages[[2]][,,1], pancreasMasks[[2]], sd)[-1],
tapply(pancreasImages[[3]][,,1], pancreasMasks[[3]], sd)[-1])
cur_H3[is.na(cur_H3)] <- 0
cur_CD99 <- c(tapply(pancreasImages[[1]][,,2], pancreasMasks[[1]], sd)[-1],
tapply(pancreasImages[[2]][,,2], pancreasMasks[[2]], sd)[-1],
tapply(pancreasImages[[3]][,,2], pancreasMasks[[3]], sd)[-1])
cur_CD99[is.na(cur_CD99)] <- 0
cur_PIN <- c(tapply(pancreasImages[[1]][,,3], pancreasMasks[[1]], sd)[-1],
tapply(pancreasImages[[2]][,,3], pancreasMasks[[2]], sd)[-1],
tapply(pancreasImages[[3]][,,3], pancreasMasks[[3]], sd)[-1])
cur_PIN[is.na(cur_PIN)] <- 0
cur_CD8a <- c(tapply(pancreasImages[[1]][,,4], pancreasMasks[[1]], sd)[-1],
tapply(pancreasImages[[2]][,,4], pancreasMasks[[2]], sd)[-1],
tapply(pancreasImages[[3]][,,4], pancreasMasks[[3]], sd)[-1])
cur_CD8a[is.na(cur_CD8a)] <- 0
cur_CDH <- c(tapply(pancreasImages[[1]][,,5], pancreasMasks[[1]], sd)[-1],
tapply(pancreasImages[[2]][,,5], pancreasMasks[[2]], sd)[-1],
tapply(pancreasImages[[3]][,,5], pancreasMasks[[3]], sd)[-1])
cur_CDH[is.na(cur_CDH)] <- 0
expect_equal(assayNames(sce), "counts_sd")
# SD
expect_equal(as.numeric(cur_H3), assay(sce, "counts_sd")["H3",])
expect_equal(as.numeric(cur_CD99), assay(sce, "counts_sd")["CD99",])
expect_equal(as.numeric(cur_PIN), assay(sce, "counts_sd")["PIN",])
expect_equal(as.numeric(cur_CD8a), assay(sce, "counts_sd")["CD8a",])
expect_equal(as.numeric(cur_CDH), assay(sce, "counts_sd")["CDH",])
expect_silent(sce <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb", basic_feature = "mad"))
cur_H3 <- c(tapply(pancreasImages[[1]][,,1], pancreasMasks[[1]], mad)[-1],
tapply(pancreasImages[[2]][,,1], pancreasMasks[[2]], mad)[-1],
tapply(pancreasImages[[3]][,,1], pancreasMasks[[3]], mad)[-1])
cur_H3[is.na(cur_H3)] <- 0
cur_CD99 <- c(tapply(pancreasImages[[1]][,,2], pancreasMasks[[1]], mad)[-1],
tapply(pancreasImages[[2]][,,2], pancreasMasks[[2]], mad)[-1],
tapply(pancreasImages[[3]][,,2], pancreasMasks[[3]], mad)[-1])
cur_CD99[is.na(cur_CD99)] <- 0
cur_PIN <- c(tapply(pancreasImages[[1]][,,3], pancreasMasks[[1]], mad)[-1],
tapply(pancreasImages[[2]][,,3], pancreasMasks[[2]], mad)[-1],
tapply(pancreasImages[[3]][,,3], pancreasMasks[[3]], mad)[-1])
cur_PIN[is.na(cur_PIN)] <- 0
cur_CD8a <- c(tapply(pancreasImages[[1]][,,4], pancreasMasks[[1]], mad)[-1],
tapply(pancreasImages[[2]][,,4], pancreasMasks[[2]], mad)[-1],
tapply(pancreasImages[[3]][,,4], pancreasMasks[[3]], mad)[-1])
cur_CD8a[is.na(cur_CD8a)] <- 0
cur_CDH <- c(tapply(pancreasImages[[1]][,,5], pancreasMasks[[1]], mad)[-1],
tapply(pancreasImages[[2]][,,5], pancreasMasks[[2]], mad)[-1],
tapply(pancreasImages[[3]][,,5], pancreasMasks[[3]], mad)[-1])
cur_CDH[is.na(cur_CDH)] <- 0
expect_equal(assayNames(sce), "counts_mad")
# MAD
expect_equal(as.numeric(cur_H3), assay(sce, "counts_mad")["H3",])
expect_equal(as.numeric(cur_CD99), assay(sce, "counts_mad")["CD99",])
expect_equal(as.numeric(cur_PIN), assay(sce, "counts_mad")["PIN",])
expect_equal(as.numeric(cur_CD8a), assay(sce, "counts_mad")["CD8a",])
expect_equal(as.numeric(cur_CDH), assay(sce, "counts_mad")["CDH",])
expect_silent(sce <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb", basic_feature = "q05", basic_quantiles = 0.5))
cur_H3 <- c(tapply(pancreasImages[[1]][,,1], pancreasMasks[[1]], median)[-1],
tapply(pancreasImages[[2]][,,1], pancreasMasks[[2]], median)[-1],
tapply(pancreasImages[[3]][,,1], pancreasMasks[[3]], median)[-1])
cur_CD99 <- c(tapply(pancreasImages[[1]][,,2], pancreasMasks[[1]], median)[-1],
tapply(pancreasImages[[2]][,,2], pancreasMasks[[2]], median)[-1],
tapply(pancreasImages[[3]][,,2], pancreasMasks[[3]], median)[-1])
cur_PIN <- c(tapply(pancreasImages[[1]][,,3], pancreasMasks[[1]], median)[-1],
tapply(pancreasImages[[2]][,,3], pancreasMasks[[2]], median)[-1],
tapply(pancreasImages[[3]][,,3], pancreasMasks[[3]], median)[-1])
cur_CD8a <- c(tapply(pancreasImages[[1]][,,4], pancreasMasks[[1]], median)[-1],
tapply(pancreasImages[[2]][,,4], pancreasMasks[[2]], median)[-1],
tapply(pancreasImages[[3]][,,4], pancreasMasks[[3]], median)[-1])
cur_CDH <- c(tapply(pancreasImages[[1]][,,5], pancreasMasks[[1]], median)[-1],
tapply(pancreasImages[[2]][,,5], pancreasMasks[[2]], median)[-1],
tapply(pancreasImages[[3]][,,5], pancreasMasks[[3]], median)[-1])
expect_equal(assayNames(sce), "counts_q05")
# Median
expect_equal(as.numeric(cur_H3), assay(sce, "counts_q05")["H3",])
expect_equal(as.numeric(cur_CD99), assay(sce, "counts_q05")["CD99",])
expect_equal(as.numeric(cur_PIN), assay(sce, "counts_q05")["PIN",])
expect_equal(as.numeric(cur_CD8a), assay(sce, "counts_q05")["CD8a",])
expect_equal(as.numeric(cur_CDH), assay(sce, "counts_q05")["CDH",])
# Shape features
expect_silent(sce <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb", feature_types = c("basic", "shape"),
shape_feature = c('area', 'perimeter', 'radius.mean', 'radius.sd', 'radius.max', 'radius.min')))
expect_equal(names(colData(sce)), c("ImageNb", "object_id", "s.area", "s.perimeter", "s.radius.mean", "s.radius.sd", "s.radius.min", "s.radius.max", "ImageName"))
expect_equal(sce$s.perimeter[40:50], c(28, 39, 25, 10, 16, 15, 13, 22, 31, 37, 5))
expect_equal(sce$s.radius.sd[40:50], c(0.6381883, 1.2634757, 0.5599216, 0.5707716, 0.6687795, 0.4334581, 0.7503223, 0.4679423, 0.7259625, 0.8618819, 0.2850134), tolerance = 0.00001)
expect_equal(sce$s.radius.min[40:50], c(3.5959870, 3.1786599, 2.9772471, 0.7810250, 1.4710753, 1.2292726, 0.8770580, 2.8477191, 3.2325735, 3.9868300, 0.4472136), tolerance = 0.00001)
expect_equal(sce$s.radius.max[40:50], c(6.123933, 8.641064, 5.209990, 2.570992, 3.628231, 2.996294, 3.328201, 4.788286, 6.002514, 7.190662, 1.264911), tolerance = 0.00001)
# Moment features
expect_silent(sce <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb", feature_types = c("basic", "moment"),
moment_feature = c('cx', 'cy', 'majoraxis', 'eccentricity', 'theta')))
expect_equal(names(colData(sce)), c("ImageNb", "object_id", "m.cx", "m.cy", "m.majoraxis", "m.eccentricity", "m.theta", "ImageName"))
expect_equal(sce$m.theta[40:50], c(0.60960841, -0.04246836, -0.68045282, 1.57079633, 0.89866183, -1.11289559, -1.10737273, -0.96767979, 1.53796956, -0.09403954, -1.24904577))
# Haralick features
expect_silent(sce <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb", feature_types = c("basic", "haralick"),
haralick_feature = c("asm.s1", "con.s1", "cor.s1", "var.s1", "idm.s1", "sav.s1", "sva.s1", "sen.s1", "ent.s1", "dva.s1", "den.s1", "f12.s1", "f13.s1", "asm.s2", "con.s2", "cor.s2", "var.s2", "idm.s2", "sav.s2", "sva.s2", "sen.s2", "ent.s2", "dva.s2", "den.s2", "f12.s2", "f13.s2")))
features <- c("asm.s1", "con.s1", "cor.s1", "var.s1", "idm.s1", "sav.s1", "sva.s1", "sen.s1", "ent.s1", "dva.s1", "den.s1", "f12.s1", "f13.s1", "asm.s2", "con.s2", "cor.s2", "var.s2", "idm.s2", "sav.s2", "sva.s2", "sen.s2", "ent.s2", "dva.s2", "den.s2", "f12.s2", "f13.s2")
features <- paste0("h.", features)
features <- paste0(rep(channelNames(pancreasImages), each = length(features)), ".", features)
features <- c("ImageNb", "object_id", features, "ImageName")
expect_equal(names(colData(sce)), features)
cur_basic <- computeFeatures.basic(pancreasMasks[[1]], pancreasImages[[1]][,,1])
cur_haralick <- computeFeatures.haralick(pancreasMasks[[1]], pancreasImages[[1]][,,1])
cur_haralick <- cur_haralick[as.numeric(rownames(cur_basic)),]
colnames(cur_haralick) <- paste0("H3.", colnames(cur_haralick))
expect_equal(as.matrix(colData(sce)[sce$ImageNb == 1,colnames(cur_haralick)]), cur_haralick)
cur_haralick <- computeFeatures.haralick(pancreasMasks[[1]], pancreasImages[[1]][,,2])
cur_haralick <- cur_haralick[as.numeric(rownames(cur_basic)),]
colnames(cur_haralick) <- paste0("CD99.", colnames(cur_haralick))
expect_equal(as.matrix(colData(sce)[sce$ImageNb == 1,colnames(cur_haralick)]), cur_haralick)
expect_silent(sce <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb", feature_types = c("basic", "haralick"), haralick_scales = c(1,4),
haralick_feature = c("asm.s1", "con.s1", "cor.s1", "var.s1", "idm.s1", "sav.s1", "sva.s1", "sen.s1", "ent.s1", "dva.s1", "den.s1", "f12.s1", "f13.s1", "asm.s4", "con.s4", "cor.s4", "var.s4", "idm.s4", "sav.s4", "sva.s4", "sen.s4", "ent.s4", "dva.s4", "den.s4", "f12.s4", "f13.s4")))
features <- c("asm.s1", "con.s1", "cor.s1", "var.s1", "idm.s1", "sav.s1", "sva.s1", "sen.s1", "ent.s1", "dva.s1", "den.s1", "f12.s1", "f13.s1", "asm.s4", "con.s4", "cor.s4", "var.s4", "idm.s4", "sav.s4", "sva.s4", "sen.s4", "ent.s4", "dva.s4", "den.s4", "f12.s4", "f13.s4")
features <- paste0("h.", features)
features <- paste0(rep(channelNames(pancreasImages), each = length(features)), ".", features)
features <- c("ImageNb", "object_id", features, "ImageName")
expect_equal(names(colData(sce)), features)
cur_basic <- computeFeatures.basic(pancreasMasks[[1]], pancreasImages[[1]][,,1])
cur_haralick <- computeFeatures.haralick(pancreasMasks[[1]], pancreasImages[[1]][,,1], haralick.scales = c(1,4))
cur_haralick <- cur_haralick[as.numeric(rownames(cur_basic)),]
colnames(cur_haralick) <- paste0("H3.", colnames(cur_haralick))
expect_equal(as.matrix(colData(sce)[sce$ImageNb == 1,colnames(cur_haralick)]), cur_haralick)
cur_haralick <- computeFeatures.haralick(pancreasMasks[[1]], pancreasImages[[1]][,,2], haralick.scales = c(1,4))
cur_haralick <- cur_haralick[as.numeric(rownames(cur_basic)),]
colnames(cur_haralick) <- paste0("CD99.", colnames(cur_haralick))
expect_equal(as.matrix(colData(sce)[sce$ImageNb == 1,colnames(cur_haralick)]), cur_haralick)
expect_silent(sce <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb", feature_types = c("basic", "haralick"), haralick_nbins = 16,
haralick_feature = c("asm.s1", "con.s1", "cor.s1", "var.s1", "idm.s1", "sav.s1", "sva.s1", "sen.s1", "ent.s1", "dva.s1", "den.s1", "f12.s1", "f13.s1", "asm.s2", "con.s2", "cor.s2", "var.s2", "idm.s2", "sav.s2", "sva.s2", "sen.s2", "ent.s2", "dva.s2", "den.s2", "f12.s2", "f13.s2")))
features <- c("asm.s1", "con.s1", "cor.s1", "var.s1", "idm.s1", "sav.s1", "sva.s1", "sen.s1", "ent.s1", "dva.s1", "den.s1", "f12.s1", "f13.s1", "asm.s2", "con.s2", "cor.s2", "var.s2", "idm.s2", "sav.s2", "sva.s2", "sen.s2", "ent.s2", "dva.s2", "den.s2", "f12.s2", "f13.s2")
features <- paste0("h.", features)
features <- paste0(rep(channelNames(pancreasImages), each = length(features)), ".", features)
features <- c("ImageNb", "object_id", features, "ImageName")
expect_equal(names(colData(sce)), features)
cur_basic <- computeFeatures.basic(pancreasMasks[[1]], pancreasImages[[1]][,,1])
cur_haralick <- computeFeatures.haralick(pancreasMasks[[1]], pancreasImages[[1]][,,1], haralick.nbins = 16)
cur_haralick <- cur_haralick[as.numeric(rownames(cur_basic)),]
colnames(cur_haralick) <- paste0("H3.", colnames(cur_haralick))
expect_equal(as.matrix(colData(sce)[sce$ImageNb == 1,colnames(cur_haralick)]), cur_haralick)
cur_haralick <- computeFeatures.haralick(pancreasMasks[[1]], pancreasImages[[1]][,,2], haralick.nbins = 16)
cur_haralick <- cur_haralick[as.numeric(rownames(cur_basic)),]
colnames(cur_haralick) <- paste0("CD99.", colnames(cur_haralick))
expect_equal(as.matrix(colData(sce)[sce$ImageNb == 1,colnames(cur_haralick)]), cur_haralick)
expect_silent(sce <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb", feature_types = c("basic", "haralick"),
haralick_feature = c("asm.s1", "ent.s2")))
features <- c("asm.s1", "ent.s2")
features <- paste0("h.", features)
features <- paste0(rep(channelNames(pancreasImages), each = length(features)), ".", features)
features <- c("ImageNb", "object_id", features, "ImageName")
expect_equal(names(colData(sce)), features)
cur_basic <- computeFeatures.basic(pancreasMasks[[1]], pancreasImages[[1]][,,1])
cur_haralick <- computeFeatures.haralick(pancreasMasks[[1]], pancreasImages[[1]][,,1])
cur_haralick <- cur_haralick[as.numeric(rownames(cur_basic)),]
colnames(cur_haralick) <- paste0("H3.", colnames(cur_haralick))
expect_equal(as.matrix(colData(sce)[sce$ImageNb == 1,c("H3.h.asm.s1", "H3.h.ent.s2")]), cur_haralick[,c("H3.h.asm.s1", "H3.h.ent.s2")])
cur_haralick <- computeFeatures.haralick(pancreasMasks[[1]], pancreasImages[[1]][,,2])
cur_haralick <- cur_haralick[as.numeric(rownames(cur_basic)),]
colnames(cur_haralick) <- paste0("CD99.", colnames(cur_haralick))
expect_equal(as.matrix(colData(sce)[sce$ImageNb == 1,c("CD99.h.asm.s1", "CD99.h.ent.s2")]), cur_haralick[,c("CD99.h.asm.s1", "CD99.h.ent.s2")])
})
test_that("measureObjects: defaults work", {
data("pancreasSCE")
data("pancreasMasks")
data("pancreasImages")
# Works
expect_silent(spe <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb", return_as = "spe"))
cur_H3 <- c(tapply(pancreasImages[[1]][,,1], pancreasMasks[[1]], mean)[-1],
tapply(pancreasImages[[2]][,,1], pancreasMasks[[2]], mean)[-1],
tapply(pancreasImages[[3]][,,1], pancreasMasks[[3]], mean)[-1])
cur_CD99 <- c(tapply(pancreasImages[[1]][,,2], pancreasMasks[[1]], mean)[-1],
tapply(pancreasImages[[2]][,,2], pancreasMasks[[2]], mean)[-1],
tapply(pancreasImages[[3]][,,2], pancreasMasks[[3]], mean)[-1])
cur_PIN <- c(tapply(pancreasImages[[1]][,,3], pancreasMasks[[1]], mean)[-1],
tapply(pancreasImages[[2]][,,3], pancreasMasks[[2]], mean)[-1],
tapply(pancreasImages[[3]][,,3], pancreasMasks[[3]], mean)[-1])
cur_CD8a <- c(tapply(pancreasImages[[1]][,,4], pancreasMasks[[1]], mean)[-1],
tapply(pancreasImages[[2]][,,4], pancreasMasks[[2]], mean)[-1],
tapply(pancreasImages[[3]][,,4], pancreasMasks[[3]], mean)[-1])
cur_CDH <- c(tapply(pancreasImages[[1]][,,5], pancreasMasks[[1]], mean)[-1],
tapply(pancreasImages[[2]][,,5], pancreasMasks[[2]], mean)[-1],
tapply(pancreasImages[[3]][,,5], pancreasMasks[[3]], mean)[-1])
expect_equal(assayNames(spe), "counts")
# Intensities
expect_equal(as.numeric(cur_H3), counts(spe)["H3",])
expect_equal(as.numeric(cur_CD99), counts(spe)["CD99",])
expect_equal(as.numeric(cur_PIN), counts(spe)["PIN",])
expect_equal(as.numeric(cur_CD8a), counts(spe)["CD8a",])
expect_equal(as.numeric(cur_CDH), counts(spe)["CDH",])
# Other entries
expect_equal(spe$ImageNb, pancreasSCE$ImageNb)
expect_equal(spe$object_id, pancreasSCE$CellNb)
expect_equal(rownames(spe), rownames(pancreasSCE))
expect_equal(rownames(spe), channelNames(pancreasImages))
# colData entries
expect_equal(names(colData(spe)),
c("ImageNb", "object_id", "s.area", "s.radius.mean",
"m.majoraxis", "m.eccentricity", "sample_id", "ImageName"))
area <- c(table(pancreasMasks[[1]])[-1],
table(pancreasMasks[[2]])[-1],
table(pancreasMasks[[3]])[-1])
expect_equal(spe$s.area, as.numeric(area))
expect_equal(spe$s.radius.mean[20:30],
c(4.636644, 4.764700, 3.166851, 3.637420, 4.538360, 4.423855, 5.031715, 4.556145, 4.854957, 2.911706, 3.895167),
tolerance = 0.0001)
expect_equal(SpatialExperiment::spatialCoords(spe)[20:30,"m.cx"],
c(3.776316, 29.804878, 97.243902, 58.478261, 80.092105, 39.873239, 70.260870, 12.092105, 84.920000, 2.588235, 21.271186),
tolerance = 0.0001)
expect_equal(SpatialExperiment::spatialCoords(spe)[20:30,"m.cy"],
c(13.71053, 17.71951, 16.12195, 18.56522, 18.36842, 21.67606, 23.11957, 23.78947, 25.02667, 24.50000, 27.27119),
tolerance = 0.0001)
expect_equal(spe$m.majoraxis[20:30],
c(14.651483, 12.792813, 8.718651, 10.983645, 11.665788, 11.333186, 13.358161, 13.151296, 16.216661, 9.080632, 9.137610),
tolerance = 0.0001)
expect_equal(spe$m.eccentricity[20:30],
c(0.8771370, 0.7160534, 0.7079446, 0.8471241, 0.6798998, 0.6019294, 0.7431450, 0.8167983, 0.9096242, 0.8334293, 0.4357446),
tolerance = 0.0001)
# Other standard input
spe <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb",
feature_types = "basic", return_as = "spe")
expect_equal(names(colData(spe)),
c("ImageNb", "object_id", "sample_id", "ImageName"))
spe <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb",
feature_types = c("basic", "shape"), return_as = "spe")
expect_equal(names(colData(spe)),
c("ImageNb", "object_id", "s.area", "s.radius.mean", "sample_id", "ImageName"))
spe <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb",
feature_types = c("basic", "shape", "moment"), return_as = "spe")
expect_equal(names(colData(spe)),
c("ImageNb", "object_id", "s.area", "s.radius.mean",
"m.majoraxis", "m.eccentricity", "sample_id", "ImageName"))
expect_equal(colnames(SpatialExperiment::spatialCoords(spe)),
c("m.cx", "m.cy"))
spe <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb",
feature_types = c("basic", "shape", "moment", "haralick"),
haralick_feature = "ent.s1", return_as = "spe")
expect_equal(names(colData(spe)),
c("ImageNb", "object_id", "s.area", "s.radius.mean",
"m.majoraxis", "m.eccentricity",
"H3.h.ent.s1", "CD99.h.ent.s1", "PIN.h.ent.s1", "CD8a.h.ent.s1", "CDH.h.ent.s1",
"sample_id", "ImageName"))
expect_equal(spe$H3.h.ent.s1[20:30],
c(0.92239897, 0.18210610, 0.00000000, 0.28466019, 0.24327757, 0.28306235, 0.31983314, 0.26878196, 0.05829775, 0.00000000, 0.54267473),
tolerance = 0.0001)
expect_equal(spe$CD99.h.ent.s1[20:30],
c(1.0087171, 0.6879652, 0.0000000, 0.8714867, 0.8615940, 0.6785749, 0.8175975, 0.9346840, 0.7556055, 0.8351088, 0.7464104),
tolerance = 0.0001)
expect_equal(spe$PIN.h.ent.s1[20:30],
c(0.3824506, 0.4245154, 0.7619950, 0.5764059, 0.6587099, 0.5013036, 0.4560755, 0.1652518, 0.7064710, 0.2808083, 0.6687518),
tolerance = 0.0001)
expect_equal(spe$CD8a.h.ent.s1[20:30],
c(0.4807238, 0.3278883, 0.4608923, 0.5163900, 0.4407971, 0.4067465, 0.4099984, 0.5272945, 0.5660128, 0.6652107, 0.4108133),
tolerance = 0.0001)
expect_equal(spe$CDH.h.ent.s1[20:30],
c(0.2464482, 0.9223040, 0.3924199, 0.6494582, 0.7329984, 0.9315071, 0.5924960, 0.7332656, 0.6537887, 0.6607878, 0.7881618),
tolerance = 0.0001)
spe <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb",
feature_types = c("basic", "moment"), return_as = "spe")
expect_equal(names(colData(spe)),
c("ImageNb", "object_id",
"m.majoraxis", "m.eccentricity", "sample_id", "ImageName"))
spe <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb",
feature_types = c("moment", "basic"), return_as = "spe")
expect_equal(names(colData(spe)),
c("ImageNb", "object_id",
"m.majoraxis", "m.eccentricity", "sample_id", "ImageName"))
spe <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb",
feature_types = c("basic", "haralick"),
haralick_feature = "ent.s1", return_as = "spe")
expect_equal(names(colData(spe)),
c("ImageNb", "object_id",
"H3.h.ent.s1", "CD99.h.ent.s1", "PIN.h.ent.s1", "CD8a.h.ent.s1", "CDH.h.ent.s1", "sample_id", "ImageName"))
# Parallelisable
expect_silent(spe.1 <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb", return_as = "spe",
BPPARAM = BiocParallel::bpparam()))
expect_silent(spe.2 <- measureObjects(pancreasMasks, pancreasImages, img_id = "ImageNb", return_as = "spe",
BPPARAM = BiocParallel::SerialParam()))
expect_equal(spe.1, spe.2)
expect_equal(as.numeric(cur_H3), counts(spe.1)["H3",])
# different mcols entries work
cur_images <- pancreasImages
mcols(cur_images)$other_entry <- "test"
spe <- measureObjects(pancreasMasks, cur_images, img_id = "ImageNb", return_as = "spe")
expect_equal(names(colData(spe)),
c("ImageNb", "object_id", "s.area", "s.radius.mean",
"m.majoraxis", "m.eccentricity", "sample_id", "ImageName",
"other_entry"))
cur_masks <- pancreasMasks
mcols(cur_masks)$other_entry_2 <- c("test1", "test2", "test3")
spe <- measureObjects(cur_masks, cur_images, img_id = "ImageNb", return_as = "spe")
expect_equal(names(colData(spe)),
c("ImageNb", "object_id", "s.area", "s.radius.mean",
"m.majoraxis", "m.eccentricity", "sample_id", "ImageName",
"other_entry", "other_entry_2"))
})
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Embedding an R snippet on your website
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.