The NCBI recently started adding NA values as FTP file paths in the…

…ir `species summary files` for species without reference genomes. As a result `meta.retrieval()` stopped working, because no FTP paths were found for some species. This issue was now fixed by adding the filter rule `!is.na(ftp_path)` into all `get*()` functions (Many thanks for making me aware of this issue Ashok Kumar Sharma #34 and Dominik Merges #72)
ropensci · Mar 17, 2021 · ac025b9 · ac025b9
1 parent 092bf7a
commit ac025b9
Show file tree

Hide file tree

Showing 7 changed files with 32 additions and 32 deletions.
diff --git a/R/getAssemblyStats.R b/R/getAssemblyStats.R
@@ -139,14 +139,14 @@ getAssemblyStats <-
                         AssemblyFilesAllKingdoms,
                         stringr::str_detect(organism_name, organism) |
                             stringr::str_detect(assembly_accession, organism),
-                        (version_status == "latest")
+                        (version_status == "latest"), !is.na(ftp_path)
                     ) 
             } else {
                 FoundOrganism <-
                     dplyr::filter(
                         AssemblyFilesAllKingdoms,
                         taxid == as.integer(organism),
-                        (version_status == "latest")
+                        (version_status == "latest"), !is.na(ftp_path)
                     ) 
             }
         }
@@ -323,7 +323,7 @@ getAssemblyStats <-
 
                 )
 
-                readr::write_tsv(doc, path = file.path(path,paste0("doc_",local.org,"_db_",db,".tsv")))
+                readr::write_tsv(doc, file = file.path(path,paste0("doc_",local.org,"_db_",db,".tsv")))
 
                 message(
                     paste0(

diff --git a/R/getCDS.R b/R/getCDS.R
@@ -119,14 +119,14 @@ getCDS <-
                             AssemblyFilesAllKingdoms,
                             stringr::str_detect(organism_name, organism) |
                                 stringr::str_detect(assembly_accession, organism),
-                            (version_status == "latest")
+                            (version_status == "latest"), !is.na(ftp_path)
                         ) 
                 } else {
                     FoundOrganism <-
                         dplyr::filter(
                             AssemblyFilesAllKingdoms,
                             taxid == as.integer(organism),
-                            (version_status == "latest")
+                            (version_status == "latest"), !is.na(ftp_path)
                         ) 
                 }
             }
@@ -322,7 +322,7 @@ getCDS <-
 
                     )
 
-                    readr::write_tsv(doc, path = file.path(path,paste0("doc_",local.org,"_db_",db,".tsv")))
+                    readr::write_tsv(doc, file = file.path(path,paste0("doc_",local.org,"_db_",db,".tsv")))
 
                     message(
                         paste0(
@@ -504,7 +504,7 @@ getCDS <-
 
                 )
 
-                readr::write_tsv(doc, file.path(
+                readr::write_tsv(doc, file = file.path(
                     path,
                     paste0("doc_", new.organism, "_db_", db, ".tsv"))
                 )
@@ -672,7 +672,7 @@ getCDS <-
 
                 )
 
-                readr::write_tsv(doc, file.path(
+                readr::write_tsv(doc, file = file.path(
                     path,
                     paste0("doc_", new.organism, "_db_", db, ".tsv"))
                 )

diff --git a/R/getGFF.R b/R/getGFF.R
@@ -134,14 +134,14 @@ getGFF <-
                             AssemblyFilesAllKingdoms,
                             stringr::str_detect(organism_name, organism) |
                                 stringr::str_detect(assembly_accession, organism),
-                            (version_status == "latest")
+                            (version_status == "latest"), !is.na(ftp_path)
                         ) 
                 } else {
                     FoundOrganism <-
                         dplyr::filter(
                             AssemblyFilesAllKingdoms,
                             taxid == as.integer(organism),
-                            (version_status == "latest")
+                            (version_status == "latest"), !is.na(ftp_path)
                         ) 
                 }
             }
@@ -322,7 +322,7 @@ getGFF <-
 
                     )
 
-                    readr::write_tsv(doc, path = file.path(path,paste0("doc_",local.org,"_db_",db,".tsv")))
+                    readr::write_tsv(doc, file = file.path(path,paste0("doc_",local.org,"_db_",db,".tsv")))
 
                     if (gunzip) {
                             message(
@@ -488,7 +488,7 @@ getGFF <-
 
                 )
 
-                readr::write_tsv(doc, file.path(
+                readr::write_tsv(doc, file = file.path(
                     path,
                     paste0("doc_", new.organism, "_db_", db, ".tsv"))
                 )
@@ -659,7 +659,7 @@ getGFF <-
 
                 )
 
-                readr::write_tsv(doc, file.path(
+                readr::write_tsv(doc, file = file.path(
                     path,
                     paste0("doc_", new.organism, "_db_", db, ".tsv"))
                 )

diff --git a/R/getGenome.R b/R/getGenome.R
@@ -161,14 +161,14 @@ getGenome <-
                             AssemblyFilesAllKingdoms,
                             stringr::str_detect(organism_name, organism) |
                                 assembly_accession == organism,
-                            (version_status == "latest")
+                            (version_status == "latest"), !is.na(ftp_path)
                         )
                 } else {
                     FoundOrganism <-
                         dplyr::filter(
                             AssemblyFilesAllKingdoms,
                             taxid == as.integer(organism),
-                            (version_status == "latest")
+                            (version_status == "latest"), !is.na(ftp_path)
                         )
                 }
             }
@@ -339,12 +339,12 @@ getGenome <-
 
                     )
 
-                    readr::write_tsv(doc, path = file.path(path, paste0("doc_",local.org,"_db_",db,".tsv")))
+                    readr::write_tsv(doc, file = file.path(path, paste0("doc_",local.org,"_db_",db,".tsv")))
 
                     genome_summary_stats <- summary_genome(file = file.path(path,
                                                                             paste0(local.org, "_genomic_", db, ".fna.gz")), organism = organism)
 
-                    readr::write_tsv(genome_summary_stats, path = file.path(path, paste0("doc_",local.org,"_db_",db,"_summary_statistics.tsv")))
+                    readr::write_tsv(genome_summary_stats, file = file.path(path, paste0("doc_",local.org,"_db_",db,"_summary_statistics.tsv")))
 
                     if (!gunzip) {
                             message(
@@ -529,7 +529,7 @@ getGenome <-
 
                 )
 
-                readr::write_tsv(doc, file.path(
+                readr::write_tsv(doc, file = file.path(
                         path,
                         paste0("doc_", new.organism, "_db_", db, ".tsv"))
                         )
@@ -696,7 +696,7 @@ getGenome <-
 
                 )
 
-                readr::write_tsv(doc, file.path(
+                readr::write_tsv(doc, file = file.path(
                         path,
                         paste0("doc_", new.organism, "_db_", db, ".tsv"))
                 )

diff --git a/R/getProteome.R b/R/getProteome.R
@@ -141,14 +141,14 @@ getProteome <-
                             AssemblyFilesAllKingdoms,
                             stringr::str_detect(organism_name, organism) |
                                 stringr::str_detect(assembly_accession, organism),
-                            (version_status == "latest")
+                            (version_status == "latest"), !is.na(ftp_path)
                         ) 
                 } else {
                     FoundOrganism <-
                         dplyr::filter(
                             AssemblyFilesAllKingdoms,
                             taxid == as.integer(organism),
-                            (version_status == "latest")
+                            (version_status == "latest"), !is.na(ftp_path)
                         ) 
                 }
             }
@@ -325,7 +325,7 @@ getProteome <-
 
                     )
 
-                    readr::write_tsv(doc, path = file.path(path,paste0("doc_",local.org,"_db_",db,".tsv")))
+                    readr::write_tsv(doc, file = file.path(path,paste0("doc_",local.org,"_db_",db,".tsv")))
 
                     if (!gunzip) {
                             message(
@@ -506,7 +506,7 @@ getProteome <-
 
                 )
 
-                readr::write_tsv(doc, file.path(
+                readr::write_tsv(doc, file = file.path(
                         path,
                         paste0("doc_", new.organism, "_db_", db, ".tsv"))
                 )
@@ -673,7 +673,7 @@ getProteome <-
 
                 )
 
-                readr::write_tsv(doc, file.path(
+                readr::write_tsv(doc, file = file.path(
                         path,
                         paste0("doc_", new.organism, "_db_", db, ".tsv"))
                 )

diff --git a/R/getRNA.R b/R/getRNA.R
@@ -115,14 +115,14 @@ getRNA <-
                             AssemblyFilesAllKingdoms,
                             stringr::str_detect(organism_name, organism) |
                                 stringr::str_detect(assembly_accession, organism),
-                            (version_status == "latest")
+                            (version_status == "latest"), !is.na(ftp_path)
                         ) 
                 } else {
                     FoundOrganism <-
                         dplyr::filter(
                             AssemblyFilesAllKingdoms,
                             taxid == as.integer(organism),
-                            (version_status == "latest")
+                            (version_status == "latest"), !is.na(ftp_path)
                         ) 
                 }
             }
@@ -319,7 +319,7 @@ getRNA <-
 
                     )
 
-                    readr::write_tsv(doc, path = file.path(path,paste0("doc_",local.org,"_db_",db,".tsv")))
+                    readr::write_tsv(doc, file = file.path(path,paste0("doc_",local.org,"_db_",db,".tsv")))
 
                     message(
                         paste0(
@@ -483,7 +483,7 @@ getRNA <-
 
                 )
 
-                readr::write_tsv(doc, file.path(
+                readr::write_tsv(doc, file = file.path(
                     path,
                     paste0("doc_", new.organism, "_db_", db, ".tsv"))
                 )
@@ -628,7 +628,7 @@ getRNA <-
 
                 )
 
-                readr::write_tsv(doc, file.path(
+                readr::write_tsv(doc, file = file.path(
                     path,
                     paste0("doc_", new.organism, "_db_", db, ".tsv"))
                 )

diff --git a/R/getRepeatMasker.R b/R/getRepeatMasker.R
@@ -123,14 +123,14 @@ getRepeatMasker <-
                             AssemblyFilesAllKingdoms,
                             stringr::str_detect(organism_name, organism) |
                                 stringr::str_detect(assembly_accession, organism),
-                            (version_status == "latest")
+                            (version_status == "latest"), !is.na(ftp_path)
                         ) 
                 } else {
                     FoundOrganism <-
                         dplyr::filter(
                             AssemblyFilesAllKingdoms,
                             taxid == as.integer(organism),
-                            (version_status == "latest")
+                            (version_status == "latest"), !is.na(ftp_path)
                         ) 
                 }
             }
@@ -308,7 +308,7 @@ getRepeatMasker <-
 
                     )
 
-                    readr::write_tsv(doc, path = file.path(path,paste0("doc_",local.org,"_db_",db,".tsv")))
+                    readr::write_tsv(doc, file = file.path(path,paste0("doc_",local.org,"_db_",db,".tsv")))
 
                     message(
                         paste0(