Set class instead of omim_type on read_omim() output

To allow for generic functions downstream. Tests: PASS
DiseaseOntology · Nov 27, 2023 · 95af35a · 95af35a
1 parent 53e1b8b
commit 95af35a
Show file tree

Hide file tree

Showing 4 changed files with 30 additions and 26 deletions.
diff --git a/NEWS.md b/NEWS.md
@@ -1,5 +1,13 @@
 # DO.utils (development version)
 
+## DO Management & Analysis
+* `read_omim()` now additionally parses official OMIM downloads of search
+results and [phenotypic series titles](https://www.omim.org/phenotypicSeriesTitles/all).
+    - Includes `omim_official` attribute to indicate if the source was an
+    official download.
+    - If input is an official source, the output class will indicate the type.
+
+
 # DO.utils 0.3.0
 
 ## General

diff --git a/R/read.R b/R/read.R
@@ -147,31 +147,32 @@ read_ga <- function(ga_file, read_all = FALSE, tidy = TRUE, keep_total = FALSE,
 #' @export
 read_omim <- function(file, ...) {
     df <- preprocess_omim_dl(file, ...)
+    omim_type <- class(df)[1]
 
-    if (attr(df, "omim_type") == "search") {
+    if (omim_type == "omim_search") {
         df <- df %>%
             dplyr::mutate(
-                mim_symbol = stringr::str_extract(mim_number, "^ *[*+#%^]"),
+                mim_symbol = stringr::str_extract(.data$mim_number, "^ *[*+#%^]"),
                 mim_type = dplyr::case_match(
-                    mim_symbol,
+                    .data$mim_symbol,
                     "*" ~ "gene",
                     "+" ~ "gene, includes phenotype",
                     "#" ~ "phenotype",
                     "%" ~ "phenotype, unknown molecular basis",
                     "^" ~ "deprecated",
                     .default = "phenotype, suspected/overlap"
                 ),
-                mim_number = stringr::str_remove(mim_number, "^ *[*+#%^]"),
-                omim = paste0("OMIM:", mim_number)
+                mim_number = stringr::str_remove(.data$mim_number, "^ *[*+#%^]"),
+                omim = paste0("OMIM:", .data$mim_number)
             ) %>%
-            dplyr::relocate(omim, mim_symbol, mim_type, .before = 1)
+            dplyr::relocate("omim", "mim_symbol", "mim_type", .before = 1)
     }
 
     entry_col_ordered <- c(
         "location", "phenotype", "phenotype_mim_number", "inheritance",
         "phenotype_mapping_key", "gene_locus", "gene_locus_mim_number"
     )
-    if (attr(df, "omim_type") == "PS" || all(entry_col_ordered %in% names(df))) {
+    if (omim_type == "omim_PS" || all(entry_col_ordered %in% names(df))) {
         df <- df %>%
             dplyr::mutate(
                 omim = paste0("OMIM:", .data$phenotype_mim_number),
@@ -201,13 +202,14 @@ read_omim <- function(file, ...) {
         )
     }
 
-    if (attr(df, "omim_type") == "PS_titles") {
+    if (omim_type == "omim_PS_titles") {
         df <- df %>%
-            dplyr::mutate(omim = paste0("OMIM:", phenotypic_series_number)) %>%
-            dplyr::relocate(omim, .before = 1)
+            dplyr::mutate(
+                omim = paste0("OMIM:", .data$phenotypic_series_number)
+            ) %>%
+            dplyr::relocate("omim", .before = 1)
     }
 
-    class(df) <- c("omim_tbl", class(df))
     df
 }
 

diff --git a/R/read_helpers.R b/R/read_helpers.R
@@ -102,9 +102,9 @@ preprocess_omim_dl <- function(file, ...) {
         stringr::str_to_lower()
 
     if (!is.na(dl_type)) {
-        attr(df, "omim_type") <- dl_type
+        class(df) <- c(paste0("omim_", dl_type), "omim_tbl", class(df))
     } else {
-        attr(df, "omim_type") <- "generic"
+        class(df) <- c("omim_tbl", class(df))
     }
 
     df

diff --git a/tests/testthat/test-read.R b/tests/testthat/test-read.R
@@ -59,9 +59,8 @@ test_that("read_omim() works for OFFICIAL download of SEARCH results", {
             )
         ),
         row.names = c(NA, -14L),
-        class = c("omim_tbl", "tbl_df", "tbl", "data.frame"),
-        omim_official = TRUE,
-        omim_type = "search"
+        class = c("omim_search", "omim_tbl", "tbl_df", "tbl", "data.frame"),
+        omim_official = TRUE
     )
 
     expect_equal(read_omim("data/omim/omim-search_dl.tsv"), expected)
@@ -82,9 +81,8 @@ test_that("read_omim() works for OFFICIAL download of PHENOTYPIC SERIES TITLES",
             )
         ),
         row.names = c(NA, -4L),
-        class = c("omim_tbl", "tbl_df", "tbl", "data.frame"),
-        omim_official = TRUE,
-        omim_type = "PS_titles"
+        class = c("omim_PS_titles", "omim_tbl", "tbl_df", "tbl", "data.frame"),
+        omim_official = TRUE
     )
 
     expect_equal(read_omim("data/omim/omim-ps_titles_dl.tsv"), expected)
@@ -111,12 +109,8 @@ ps_df <- structure(
         geno_inheritance = c(NA, rep("autosomal recessive inheritance", 3))
     ),
     row.names = c(NA, -4L),
-    class = c(
-        "omim_tbl", "tbl_df", "tbl",
-        "data.frame"
-    ),
-    omim_official = TRUE,
-    omim_type = "PS"
+    class = c("omim_PS", "omim_tbl", "tbl_df", "tbl", "data.frame"),
+    omim_official = TRUE
 )
 
 test_that("read_omim() works for OFFICIAL download of PHENOTYPIC SERIES", {
@@ -128,7 +122,7 @@ test_that("read_omim() works for COPIED data (PS or with entry info)", {
         dplyr::filter(!stringr::str_detect(.data$omim, "PS")) %>%
         dplyr::mutate(gene_locus = stringr::str_remove(.data$gene_locus, ",.*"))
     attr(ps_df_cp, "omim_official") <- FALSE
-    attr(ps_df_cp, "omim_type") <- "generic"
+    class(ps_df_cp) <- class(ps_df)[-1]
 
     expect_equal(read_omim("data/omim/omim-ps_cp-ps_page.csv"), ps_df_cp)
     # expect_snapshot(read_omim("data/omim/omim-ps_cp-ps_page_w_ps.csv")) # not supported