Running ks test on multiple groups in R-CodePudding

This is my data frame a subset of my big one as an example

dput(eee)
structure(list(interactome = c("HINT-binary", "HINT-binary", 
"HINT-binary", "HINT-binary", "HINT-binary", "HINT-binary", "HINT-comp", 
"HINT-comp", "HINT-comp", "HINT-comp", "HINT-comp", "HINT-comp", 
"InBioMap", "InBioMap", "InBioMap", "InBioMap", "InBioMap", "InBioMap", 
"Menche-2015", "Menche-2015", "Menche-2015", "Menche-2015", "Menche-2015", 
"Menche-2015"), class = c("observed", "rewired", "rewired", "rewired", 
"rewired", "rewired", "observed", "rewired", "rewired", "rewired", 
"rewired", "rewired", "observed", "rewired", "rewired", "rewired", 
"rewired", "rewired", "observed", "rewired", "rewired", "rewired", 
"rewired", "rewired"), PPI = c(844L, 609L, 591L, 593L, 590L, 
608L, 1329L, 874L, 872L, 864L, 807L, 855L, 7077L, 5049L, 5051L, 
5025L, 4975L, 5014L, 2445L, 1673L, 1652L, 1716L, 1712L, 1683L
), LCC = c(290L, 191L, 188L, 214L, 183L, 215L, 401L, 346L, 365L, 
366L, 359L, 356L, 635L, 615L, 613L, 613L, 617L, 615L, 528L, 476L, 
493L, 490L, 492L, 480L)), row.names = c(1L, 2L, 3L, 4L, 5L, 6L, 
1002L, 1003L, 1004L, 1005L, 1006L, 1007L, 2003L, 2004L, 2005L, 
2006L, 2007L, 2008L, 3004L, 3005L, 3006L, 3007L, 3008L, 3009L
), class = "data.frame")

I would like to run ks test on my different groups.

My groups in the data-frame as such "HINT-binary" "HINT-comp" "InBioMap" "Menche-2015"

Here I found one this is what I'm trying to replicate KS test

The description for the figure give as such

(D) Number of protein-protein interactions (PPIs) between LC genes observed in the high-confidence human interactome (Menche et al., 2015) (dotted line) and 1000 randomized interactome networks (density), revealing significant enrichment for PPIs between LC genes relative to random expectation (p < 10−3). (E) Size of the largest connected component (LCC) between LC genes in the high-confidence human interactome (dotted line) and 1000 randomized interactome networks (density), revealing LC genes occupy a distinct region of the human interactome (p < 10−3). (F) LC genes are prioritized by a disease gene prediction algorithm (Ghiassian et al., 2015) (p < 10−15, Kolmogorov–Smirnov test).

CodePudding user response：

Consider combn to pass pairwise combinations of those groups into ks.test method:

# BUILD NESTED LIST OF RESULTS
ks_results <- combn(
  unique(eee$interactome), 
  2, 
  FUN = \(x) list(
    PPI_ks_results = ks.test(
      eee$PPI[eee$interactome == x[1]], eee$PPI[eee$interactome == x[2]]
    ),
    LCC_ks_results = ks.test(
      eee$LCC[eee$interactome == x[1]], eee$LCC[eee$interactome == x[2]]
    )
  ),
  simplify = FALSE
)

# NAME LIST ELEMENTS
ks_results_names <- setNames(
  ks_results,
  combn(
    unique(eee$interactome), 2, simplify = FALSE
  ) |> lapply(
    \(x) paste(x, collapse="_")
  )
)

Output

# REVIEW LIST AND ELEMENTS
str(ks_results)

# List of 6
# $ HINT-binary_HINT-comp  :List of 2
# ..$ PPI_ks_results:List of 5
# .. ..$ statistic  : Named num 0.833
# .. .. ..- attr(*, "names")= chr "D"
# .. ..$ p.value    : num 0.026
# .. ..$ alternative: chr "two-sided"
# .. ..$ method     : chr "Two-sample Kolmogorov-Smirnov test"
# .. ..$ data.name  : chr "eee$PPI[eee$interactome == x[1]] and eee$PPI[eee$interactome == x[2]]"
# .. ..- attr(*, "class")= chr "htest"
# ..$ LCC_ks_results:List of 5
# .. ..$ statistic  : Named num 1
# .. .. ..- attr(*, "names")= chr "D"
# .. ..$ p.value    : num 0.00216
# .. ..$ alternative: chr "two-sided"
# .. ..$ method     : chr "Two-sample Kolmogorov-Smirnov test"
# .. ..$ data.name  : chr "eee$LCC[eee$interactome == x[1]] and eee$LCC[eee$interactome == x[2]]"
# .. ..- attr(*, "class")= chr "htest"
# $ HINT-binary_InBioMap   :List of 2
# ..$ PPI_ks_results:List of 5
# .. ..$ statistic  : Named num 1
# .. .. ..- attr(*, "names")= chr "D"
# .. ..$ p.value    : num 0.00216
# .. ..$ alternative: chr "two-sided"
# .. ..$ method     : chr "Two-sample Kolmogorov-Smirnov test"
# .. ..$ data.name  : chr "eee$PPI[eee$interactome == x[1]] and eee$PPI[eee$interactome == x[2]]"
# .. ..- attr(*, "class")= chr "htest"
# ..$ LCC_ks_results:List of 5
# .. ..$ statistic  : Named num 1
# .. .. ..- attr(*, "names")= chr "D"
# .. ..$ p.value    : num 0.00496
# .. ..$ alternative: chr "two-sided"
# .. ..$ method     : chr "Two-sample Kolmogorov-Smirnov test"
# .. ..$ data.name  : chr "eee$LCC[eee$interactome == x[1]] and eee$LCC[eee$interactome == x[2]]"
# .. ..- attr(*, "class")= chr "htest"
# $ HINT-binary_Menche-2015:List of 2
# ..$ PPI_ks_results:List of 5
# .. ..$ statistic  : Named num 1
# .. .. ..- attr(*, "names")= chr "D"
# .. ..$ p.value    : num 0.00216
# .. ..$ alternative: chr "two-sided"
# .. ..$ method     : chr "Two-sample Kolmogorov-Smirnov test"
# .. ..$ data.name  : chr "eee$PPI[eee$interactome == x[1]] and eee$PPI[eee$interactome == x[2]]"
# .. ..- attr(*, "class")= chr "htest"
# ..$ LCC_ks_results:List of 5
# .. ..$ statistic  : Named num 1
# .. .. ..- attr(*, "names")= chr "D"
# .. ..$ p.value    : num 0.00216
# .. ..$ alternative: chr "two-sided"
# .. ..$ method     : chr "Two-sample Kolmogorov-Smirnov test"
# .. ..$ data.name  : chr "eee$LCC[eee$interactome == x[1]] and eee$LCC[eee$interactome == x[2]]"
# .. ..- attr(*, "class")= chr "htest"
# $ HINT-comp_InBioMap     :List of 2
# ..$ PPI_ks_results:List of 5
# .. ..$ statistic  : Named num 1
# .. .. ..- attr(*, "names")= chr "D"
# .. ..$ p.value    : num 0.00216
# .. ..$ alternative: chr "two-sided"
# .. ..$ method     : chr "Two-sample Kolmogorov-Smirnov test"
# .. ..$ data.name  : chr "eee$PPI[eee$interactome == x[1]] and eee$PPI[eee$interactome == x[2]]"
# .. ..- attr(*, "class")= chr "htest"
# ..$ LCC_ks_results:List of 5
# .. ..$ statistic  : Named num 1
# .. .. ..- attr(*, "names")= chr "D"
# .. ..$ p.value    : num 0.00496
# .. ..$ alternative: chr "two-sided"
# .. ..$ method     : chr "Two-sample Kolmogorov-Smirnov test"
# .. ..$ data.name  : chr "eee$LCC[eee$interactome == x[1]] and eee$LCC[eee$interactome == x[2]]"
# .. ..- attr(*, "class")= chr "htest"
# $ HINT-comp_Menche-2015  :List of 2
# ..$ PPI_ks_results:List of 5
# .. ..$ statistic  : Named num 1
# .. .. ..- attr(*, "names")= chr "D"
# .. ..$ p.value    : num 0.00216
# .. ..$ alternative: chr "two-sided"
# .. ..$ method     : chr "Two-sample Kolmogorov-Smirnov test"
# .. ..$ data.name  : chr "eee$PPI[eee$interactome == x[1]] and eee$PPI[eee$interactome == x[2]]"
# .. ..- attr(*, "class")= chr "htest"
# ..$ LCC_ks_results:List of 5
# .. ..$ statistic  : Named num 1
# .. .. ..- attr(*, "names")= chr "D"
# .. ..$ p.value    : num 0.00216
# .. ..$ alternative: chr "two-sided"
# .. ..$ method     : chr "Two-sample Kolmogorov-Smirnov test"
# .. ..$ data.name  : chr "eee$LCC[eee$interactome == x[1]] and eee$LCC[eee$interactome == x[2]]"
# .. ..- attr(*, "class")= chr "htest"
# $ InBioMap_Menche-2015   :List of 2
# ..$ PPI_ks_results:List of 5
# .. ..$ statistic  : Named num 1
# .. .. ..- attr(*, "names")= chr "D"
# .. ..$ p.value    : num 0.00216
# .. ..$ alternative: chr "two-sided"
# .. ..$ method     : chr "Two-sample Kolmogorov-Smirnov test"
# .. ..$ data.name  : chr "eee$PPI[eee$interactome == x[1]] and eee$PPI[eee$interactome == x[2]]"
# .. ..- attr(*, "class")= chr "htest"
# ..$ LCC_ks_results:List of 5
# .. ..$ statistic  : Named num 1
# .. .. ..- attr(*, "names")= chr "D"
# .. ..$ p.value    : num 0.00496
# .. ..$ alternative: chr "two-sided"
# .. ..$ method     : chr "Two-sample Kolmogorov-Smirnov test"
# .. ..$ data.name  : chr "eee$LCC[eee$interactome == x[1]] and eee$LCC[eee$interactome == x[2]]"
# .. ..- attr(*, "class")= chr "htest"

Access Individual Elements

ks_results$`HINT-binary_HINT-comp`$PPI_ks_results$statistic
#         D 
# 0.8333333 
ks_results$`HINT-binary_HINT-comp`$PPI_ks_results$p.value
# [1] 0.02597403

Bind to Data Frame

data.frame(
  statistic = sapply(ks_results, \(x) x$PPI_ks_results$statistic),
  p_value = sapply(ks_results, \(x) x$PPI_ks_results$p.value),
  alternative = sapply(ks_results, \(x) x$PPI_ks_results$alternative),
  method = sapply(ks_results, \(x) x$PPI_ks_results$method)
)

#                           statistic     p_value alternative                             method
# HINT-binary_HINT-comp.D   0.8333333 0.025974026   two-sided Two-sample Kolmogorov-Smirnov test
# HINT-binary_InBioMap.D    1.0000000 0.002164502   two-sided Two-sample Kolmogorov-Smirnov test
# HINT-binary_Menche-2015.D 1.0000000 0.002164502   two-sided Two-sample Kolmogorov-Smirnov test
# HINT-comp_InBioMap.D      1.0000000 0.002164502   two-sided Two-sample Kolmogorov-Smirnov test
# HINT-comp_Menche-2015.D   1.0000000 0.002164502   two-sided Two-sample Kolmogorov-Smirnov test
# InBioMap_Menche-2015.D    1.0000000 0.002164502   two-sided Two-sample Kolmogorov-Smirnov test