Add external data from the FARMM study

.Rbuildignore

@@ -1,2 +1,3 @@
 ^.*\.Rproj$
 ^\.Rproj\.user$
+^data-raw$

DESCRIPTION

@@ -24,3 +24,5 @@ Imports:
     usedist,
     dplyr (>= 1.0.0),
     tibble (>= 3.0.0)
+Depends:
+    R (>= 2.10)

R/farmm.R

@@ -0,0 +1,43 @@
+#' Food And Resulting Microbial Metabolites (FARMM) study data
+#'
+#' Information on samples collected in the FARMM study, and a distance matrix
+#' containing Bray-Curtis distances between the microbial communities in each
+#' sample.
+#'
+#' @references Tanes C, Bittinger K, Gao Y, Friedman ES, Nessel L, Paladhi UR,
+#'   Chau L, Panfen E, Fischbach MA, Braun J, Xavier RJ, Clish CB, Li H,
+#'   Bushman FD, Lewis JD, Wu GD. Role of dietary fiber in the recovery of the
+#'   human gut microbiome and its metabolome. Cell Host Microbe. 2021 Mar
+#'   10;29(3):394-407.e5. doi: 10.1016/j.chom.2020.12.012. Epub 2021 Jan 12.
+#'   PMID: 33440171; PMCID: PMC8022197.
+#' @source \url{https://pubmed.ncbi.nlm.nih.gov/33440171/}
+#' @name farmm
+
+#' @rdname farmm
+#' @format \code{farmm_samples} is a data frame with 414 rows and 11 variables:
+#' \describe{
+#'   \item{sample_id}{unique identifiers for each sample, to match the
+#'     distances in \code{farmm_bc}}
+#'   \item{subject_id}{unique identifiers for each subject in the study}
+#'   \item{study_day}{the day of the study on which the sample was collected}
+#'   \item{diet}{the diet assigned to each subject}
+#'   \item{antibiotics}{the antibiotics status of each subject during the study.
+#'     A factor with three levels: "pre" (before antibiotic exposure),
+#'     "current" (during antibiotic exposure), and "post" (after antibiotic
+#'     exposure)}
+#'   \item{height}{the height of each subject}
+#'   \item{weight}{the weight of each subject}
+#'   \item{age}{the age of each subject}
+#'   \item{bacterial_16S_copies}{the number of bacterial 16S rRNA gene copies
+#'     per gram feces, giving an estimate of absolute bacterial abundance}
+#'   \item{num_reads}{the number of high-quality, non-host sequencing reads per
+#'     sample}
+#'   \item{host_frac}{the fraction of host reads in the shotgun metagenomic DNA
+#'     sequencing data}
+#' }
+"farmm_samples"
+
+#' @rdname farmm
+#' @format \code{farmm_bc} is a \code{dist} object containing the Bray-Curtis
+#'   distances between the 414 samples listed in \code{farmm_samples}.
+"farmm_bc"

data-raw/farmm.R

@@ -0,0 +1,53 @@
+library(tidyverse)
+library(usedist)
+
+farmm_samples <- read_tsv(
+  "data-raw/farmm_samples.tsv", show_col_types = FALSE) %>%
+  filter(SampleType %in% "Feces") %>%
+  filter(Keep) %>%
+  filter(!(study_day %in% "PS")) %>%
+  mutate(study_day = as.integer(study_day)) %>%
+  mutate(study_group = fct_relevel(study_group, "Omnivore", "Vegan", "EEN")) %>%
+  mutate(current_antibiotics = fct_recode(
+    current_antibiotics, pre = "Pre Antibiotics",
+    current = "Antibiotics Treatment", post = "Post Antibiotics")) %>%
+  mutate(current_antibiotics = fct_relevel(
+    current_antibiotics, "pre", "current", "post")) %>%
+  mutate(host_frac = host / (host + non_host)) %>%
+  mutate(new_sample_id = paste(
+    "farmm", SubjectID, sprintf("%02d", study_day), sep = "."))
+
+farmm_new_sample_ids <- farmm_samples$new_sample_id
+names(farmm_new_sample_ids) <- farmm_samples$SampleID
+
+farmm_bc <- read_tsv(
+  "data-raw/farmm_bc.tsv", show_col_types = FALSE) %>%
+  rename(sample_id = `...1`) %>%
+  column_to_rownames("sample_id") %>%
+  as.matrix() %>%
+  `[`(farmm_samples$SampleID, farmm_samples$SampleID)
+
+colnames(farmm_bc) <- farmm_new_sample_ids[colnames(farmm_bc)]
+rownames(farmm_bc) <- farmm_new_sample_ids[rownames(farmm_bc)]
+
+farmm_samples <- farmm_samples %>%
+  select(
+    sample_id = new_sample_id,
+    subject_id = SubjectID,
+    study_day,
+    diet = study_group,
+    antibiotics = current_antibiotics,
+    height,
+    weight,
+    age = Age,
+    bacterial_16S_copies = copy_num_per_gram_feces,
+    num_reads = non_host,
+    host_frac) %>%
+  arrange(diet, subject_id, study_day)
+
+farmm_bc <- farmm_bc %>%
+  as.dist() %>%
+  dist_subset(farmm_samples$sample_id)
+
+usethis::use_data(farmm_samples, overwrite = TRUE)
+usethis::use_data(farmm_bc, overwrite = TRUE)

tests/testthat/test-farmm.R

@@ -0,0 +1,14 @@
+test_that("farmm_samples data is available", {
+  expect_equal(nrow(farmm_samples), 414)
+  expect_equal(colnames(farmm_samples), c(
+    "sample_id", "subject_id", "study_day", "diet", "antibiotics", "height",
+    "weight", "age", "bacterial_16S_copies", "num_reads", "host_frac"))
+  expect_equal(levels(farmm_samples$diet), c("Omnivore", "Vegan", "EEN"))
+  expect_equal(levels(farmm_samples$antibiotics), c("pre", "current", "post"))
+})
+
+test_that("farmm_bc distance matrix matches farmm_samples data frame", {
+  expect_equal(dim(as.matrix(farmm_bc)), c(414, 414))
+  expect_equal(colnames(as.matrix(farmm_bc)), farmm_samples$sample_id)
+  expect_equal(rownames(as.matrix(farmm_bc)), farmm_samples$sample_id)
+})