@@ -120,7 +120,8 @@ importFrom(dplyr,funs)

 importFrom(dplyr,left_join)

 importFrom(dplyr,mutate_all)

 importFrom(dplyr,pull)

-importFrom(dplyr,relocate)

+importFrom(dplyr,row_number)

+importFrom(dplyr,slice)

 importFrom(edgeR,DGEList)

 importFrom(edgeR,estimateCommonDisp)

 importFrom(edgeR,estimateGLMCommonDisp)

@@ -183,7 +183,7 @@ getBarcodefromAliquot <- function(aliquot){

 #' (list of barcodes, aliquot ids, etc)

 #' @param step How many items to be evaluated per API call

 #' @param FUN function that calls the API

-splitAPICall <- function(FUN,step = 20,items){

+splitAPICall <- function(FUN, step = 20, items){

    info <- NULL

    info <- tryCatch({

         for(i in 0:(ceiling(length(items)/step) - 1)){

@@ -1283,7 +1283,8 @@ getAliquot_ids <- function(barcode){

 # getBarcodeInfo(c("TCGA-OR-A5K3-01A","C3N-00321-01"))

 # barcode is: sample_submitter_id

-#' @importFrom dplyr bind_cols

+#' @importFrom dplyr bind_cols slice row_number

+#'

 getBarcodeInfo <- function(barcode) {

   baseURL <- "https://api.gdc.cancer.gov/cases/?"

   options.pretty <- "pretty=true"

@@ -1332,7 +1333,7 @@ getBarcodeInfo <- function(barcode) {

     tryCatch({

       samples$submitter_id <-

         str_extract_all(samples$submitter_id,

-                        paste(submitter_id, collapse = "|"),

+                        paste(c(submitter_id,barcode), collapse = "|"),

                         simplify = TRUE) %>% as.character

     }, error = function(e){

       samples$submitter_id <- submitter_id

@@ -1356,7 +1357,22 @@ getBarcodeInfo <- function(barcode) {

     # this is required since the sample might not have a diagnosis

     if(!any(df$submitter_id %in% diagnoses$submitter_id)){

       diagnoses$submitter_id <- NULL

-      df <- dplyr::bind_cols(df,diagnoses)

+      # The sample migth have different sample types

+      # The diagnosis the same for each one of these samples

+      # in that case we will have a 1 to mapping and binding will

+      # not work. We need then to replicate diagnosis to each sample

+      # and not each patient

+      # Cases can be replicated with getBarcodeInfo(c("BA2691R","BA2577R","BA2748R"))

+      if(nrow(diagnoses) <  nrow(df)){

+        diagnoses <- plyr::ldply(

+          1:length(results$submitter_sample_ids),

+          .fun = function(x){

+            diagnoses[x] %>% # replicate diagnoses the number of samples

+              as.data.frame() %>%

+              dplyr::slice(rep(dplyr::row_number(), sum(results$submitter_sample_ids[[x]] %in% barcode)))})

+      }

+

+      df <- dplyr::bind_cols(df %>% as.data.frame,diagnoses %>% as.data.frame)

     } else {

       df <- left_join(df, diagnoses, by = "submitter_id")

     }

@@ -1389,7 +1405,24 @@ getBarcodeInfo <- function(barcode) {

     if(!any(df$submitter_id %in% demographic$submitter_id)){

       demographic$submitter_id <- NULL

-      df <- dplyr::bind_cols(df,demographic)

+      demographic$updated_datetime  <- NULL

+      demographic$created_datetime <- NULL

+      # The sample migth have different sample types

+      # The diagnosis the same for each one of these samples

+      # in that case we will have a 1 to mapping and binding will

+      # not work. We need then to replicate diagnosis to each sample

+      # and not each patient

+      # Cases can be replicated with getBarcodeInfo(c("BA2691R","BA2577R","BA2748R"))

+      if(nrow(demographic) <  nrow(df)){

+        demographic <- plyr::ldply(

+          1:length(results$submitter_sample_ids),

+          .fun = function(x){

+            demographic[x,] %>% # replicate diagnoses the number of samples

+              as.data.frame() %>%

+              dplyr::slice(rep(dplyr::row_number(), sum(results$submitter_sample_ids[[x]] %in% barcode)))})

+      }

+

+      df <- dplyr::bind_cols(df  %>% as.data.frame,demographic)

     } else {

       df <- left_join(df,demographic, by = "submitter_id")

     }

@@ -1408,6 +1441,16 @@ getBarcodeInfo <- function(barcode) {

                       by = "submitter_id")

     })

   } else {

+

+    if(nrow(projects.info) <  nrow(df)){

+      projects.info <- plyr::ldply(

+        1:length(results$submitter_sample_ids),

+        .fun = function(x){

+          projects.info[x,] %>% # replicate diagnoses the number of samples

+            as.data.frame() %>%

+            dplyr::slice(rep(dplyr::row_number(), sum(results$submitter_sample_ids[[x]] %in% barcode)))})

+    }

+

     df <- dplyr::bind_cols(df,projects.info)

   }

@@ -163,7 +163,7 @@

 #' @importFrom  jsonlite fromJSON

 #' @importFrom knitr kable

 #' @importFrom httr timeout

-#' @importFrom dplyr pull relocate

+#' @importFrom dplyr pull

 GDCquery <- function(project,

                      data.category,

                      data.type,

@@ -600,17 +600,18 @@ GDCquery <- function(project,

     message("ooo Check if there results for the query")

     if(nrow(results) == 0) stop("Sorry, no results were found for this query")

-    # Try ordering

+    # Try ordering (needs dplyr 1.0 - still not published)

     results <- tryCatch({

-        results %>% relocate("project") %>%

-            relocate(contains("type"), .after = project) %>%

-            relocate(contains("category"), .after = project) %>%

-            relocate(contains("experimental_strategy"), .after = project) %>%

-            relocate(contains("submitter_id"), .after = project) %>%

-            relocate(contains("sample_type"), .before = experimental_strategy) %>%

-            relocate(access,.after = last_col())  %>%

-            relocate(starts_with("analysis"), .before = access) %>%

-            relocate(contains("datetime"),.after = last_col())

+        results

+    #    results %>% relocate("project") %>%

+    #        relocate(contains("type"), .after = project) %>%

+    #        relocate(contains("category"), .after = project) %>%

+    #        relocate(contains("experimental_strategy"), .after = project) %>%

+    #        relocate(contains("submitter_id"), .after = project) %>%

+    #        relocate(contains("sample_type"), .before = experimental_strategy) %>%

+    #        relocate(access,.after = last_col())  %>%

+    #        relocate(starts_with("analysis"), .before = access) %>%

+    #        relocate(contains("datetime"),.after = last_col())

     },error = function(e){

         results

     })

@@ -145,9 +145,18 @@ test_that("IDAT files is processed", {

  #    expect_true(ncol(betas) == 1)

 })

-test_that("Prepare Samples without clinical data", {

+test_that("Prepare samples without clinical data", {

     # x <-  GDCquery_clinic(project = "TCGA-LUAD", type = "clinical")

     # x[is.na(x$diagnosis_id),]

     x <- colDataPrepare(c("TCGA-80-5608-01A","TCGA-17-Z053-01A","TCGA-78-7158-01A"))

     expect_true(nrow(x) == 3)

 })

+

+test_that("Prepare multiple samples from the same patient", {

+    # https://portal.gdc.cancer.gov/cases/d7d3de82-802d-4664-8e42-d40408b129b0?bioId=548a300f-a7eb-4dc0-b9bc-5a643ef03d5d

+    x <- colDataPrepare(c("BA2691R","BA2577R","BA2748R","BA2577D"))

+    expect_true(nrow(x) == 4)

+    expect_equal(x["BA2748R","sample_type"],"Primary Blood Derived Cancer - Bone Marrow")

+    expect_equal(x["BA2577D","sample_type"],"Recurrent Blood Derived Cancer - Bone Marrow")

+    expect_true("age_at_diagnosis" %in% colnames(x))

+})