In the mosbi
package, similarities between biclusters are
computed using different possible similarity metrics.
This vignette gives an overview about the implemented metrics.
library(mosbi)
The following similarity metrics are currently implemented:
Bray-Curtis similarity (Wikipedia)
Jaccard index (Wikipedia)
overlap coefficient (Wikipedia)
Fowlkes–Mallows index (Wikipedia)
# Bray-Curtis similarity
bray_curtis <- function(s1, s2, overlap) {
return(((2 * overlap) / (s1 + s2)))
}
# Jaccard index
jaccard <- function(s1, s2, overlap) {
return(((overlap) / (s1 + s2 - overlap)))
}
# overlap coefficient
overlap <- function(s1, s2, overlap) {
return((overlap / min(s1, s2)))
}
# Fowlkes–Mallows index
folkes_mallows <- function(s1, s2, overlap) {
tp <- choose(overlap, 2)
fp <- choose(s1 - overlap, 2)
fn <- choose(s2 - overlap, 2)
return(sqrt((tp / (tp + fp)) * (tp / (tp + fn))))
}
The behavior of the similarity metrics will be evaluated for two scenarios:
Two biclusters of the same size with an increasing overlap.
Two biclusters of different sizes (One twice as big as the other) with an increasing overlap.
# Scenario 1 - two biclusters of the same size
size1_1 <- rep(1000, 1000)
size2_1 <- rep(1000, 1000)
overlap_1 <- seq(1, 1000)
# Scenario 2 - two biclusters one of size 500, the other of size 1000
size1_2 <- rep(1000, 500)
size2_2 <- rep(500, 500)
overlap_2 <- seq(1, 500)
Two biclusters of the same size:
plot(overlap_1, bray_curtis(size1_1, size2_1, overlap_1),
col = "red", type = "l", xlab = "Overlap", ylab = "Similarity",
ylim = c(0, 1)
)
lines(overlap_1, jaccard(size1_1, size2_1, overlap_1), col = "blue")
lines(overlap_1, overlap(size1_1, size2_1, overlap_1), col = "green", lty = 2)
lines(overlap_1, folkes_mallows(size1_1, size2_1, overlap_1), col = "orange")
legend(
x = .8, legend = c("Bray-Curtis", "Jaccard", "Overlap", "Fowlkes–Mallows"),
col = c("red", "blue", "green", "orange"),
lty = 1, cex = 0.8, title = "Similarity metrics"
)
Two biclusters of different sizes:
plot(overlap_2, bray_curtis(size1_2, size2_2, overlap_2),
col = "red", type = "l", xlab = "Overlap", ylab = "Similarity",
ylim = c(0, 1)
)
lines(overlap_2, jaccard(size1_2, size2_2, overlap_2), col = "blue")
lines(overlap_2, overlap(size1_2, size2_2, overlap_2), col = "green")
lines(overlap_2, folkes_mallows(size1_2, size2_2, overlap_2), col = "orange")
legend(
x = .8, legend = c("Bray-Curtis", "Jaccard", "Overlap", "Fowlkes–Mallows"),
col = c("red", "blue", "green", "orange"),
lty = 1, cex = 0.8, title = "Similarity metrics"
)
sessionInfo()
#> R version 4.4.1 (2024-06-14)
#> Platform: x86_64-pc-linux-gnu
#> Running under: Ubuntu 24.04.1 LTS
#>
#> Matrix products: default
#> BLAS: /home/biocbuild/bbs-3.20-bioc/R/lib/libRblas.so
#> LAPACK: /usr/lib/x86_64-linux-gnu/lapack/liblapack.so.3.12.0
#>
#> locale:
#> [1] LC_CTYPE=en_US.UTF-8 LC_NUMERIC=C
#> [3] LC_TIME=en_GB 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: America/New_York
#> tzcode source: system (glibc)
#>
#> attached base packages:
#> [1] stats graphics grDevices utils datasets methods base
#>
#> other attached packages:
#> [1] mosbi_1.12.0 BiocStyle_2.34.0
#>
#> loaded via a namespace (and not attached):
#> [1] tidyr_1.3.1 generics_0.1.3 sass_0.4.9
#> [4] utf8_1.2.4 class_7.3-22 lattice_0.22-6
#> [7] digest_0.6.37 magrittr_2.0.3 evaluate_1.0.1
#> [10] grid_4.4.1 RColorBrewer_1.1-3 bookdown_0.41
#> [13] fastmap_1.2.0 jsonlite_1.8.9 tinytex_0.53
#> [16] BiocManager_1.30.25 purrr_1.0.2 fansi_1.0.6
#> [19] scales_1.3.0 modeltools_0.2-23 jquerylib_0.1.4
#> [22] cli_3.6.3 isa2_0.3.6 rlang_1.1.4
#> [25] Biobase_2.66.0 munsell_0.5.1 cachem_1.1.0
#> [28] yaml_2.3.10 tools_4.4.1 parallel_4.4.1
#> [31] biclust_2.0.3.1 dplyr_1.1.4 colorspace_2.1-1
#> [34] ggplot2_3.5.1 BiocGenerics_0.52.0 vctrs_0.6.5
#> [37] R6_2.5.1 stats4_4.4.1 lifecycle_1.0.4
#> [40] magick_2.8.5 QUBIC_1.34.0 MASS_7.3-61
#> [43] pkgconfig_2.0.3 RcppParallel_5.1.9 bslib_0.8.0
#> [46] pillar_1.9.0 gtable_0.3.6 glue_1.8.0
#> [49] Rcpp_1.0.13 tidyselect_1.2.1 xfun_0.48
#> [52] tibble_3.2.1 highr_0.11 flexclust_1.4-2
#> [55] knitr_1.48 fabia_2.52.0 igraph_2.1.1
#> [58] htmltools_0.5.8.1 rmarkdown_2.28 BH_1.84.0-0
#> [61] compiler_4.4.1 additivityTests_1.1-4.2