PR2 version 4.14.0
Radiolaria from Miguel Sandin

1 Reference

  • Pipeline PR2_rads_pipeline.pdf (link)

Comments * 67 sequences were not in GenBank at time of update. Will need to rerun the import script with the missing sequences.

2 Init

  source('../PR2_init.R', echo=FALSE)
  source('../PR2_read_google.R', echo=FALSE)

  pr2_18S <- pr2 %>% 
    filter(gene == "18S_rRNA")
  
  pr2_18S_removed <- pr2_18S %>% 
    filter(!is.na(removed_version))

3 Set up the files

  target_group =  c("Radiolaria")
  target_level = "division"

  dir_pr2_update <- "Radiolaria Miguel"
  
  pr2.env$editor <- "M. Sandin"

  full_path <- function(file_name){str_c(dir_pr2_update,"/", file_name)}

  file_pr2_update_excel <- full_path("pr2_rads_updated.xlsx")  

# create the directory for taxonomy output
  dir.create(full_path("taxo"))

4 Read the original data and reformat

4.1 Read the data

  • Number of sequences = 4911
pr2_update <- import(file_pr2_update_excel, sheet = "sequences", guess_max=200000, na=c("", "-"))
  
  str_c("Number of sequences : ", nrow(pr2_update))

4.2 Add to PR2 missing sequences from Genbank

  • Run the script script_genbank_xml.R on server

  • Run second part PR2-update-GenBank.R

Note: 67 sequences are not yet in GenBank….

  pr2_missing <-  filter(pr2_update, !(genbank_accession %in% pr2_18S$genbank_accession)) 
  DT::datatable(pr2_missing, caption = "Updated sequences that are not present in PR2 ")
  
  # Use this data frame to download new sequences and then restart from beginning
  accessions <- pr2_missing %>% 
    pull(genbank_accession)
  
  saveRDS(accessions, full_path("accessions_new.rds"))
  
  export(pr2_missing, full_path("pr2_radiolaria_not_genbank.xlsx"))

4.3 Sequences not in Genbank are removed from update

Number of sequences in Genbank : 4845

  pr2_update <- pr2_update %>% 
    filter(!(genbank_accession) %in% pr2_missing$genbank_accession) 

  str_c("Number of sequences in Genbank : ", nrow(pr2_update))

4.4 Compare with sequences in PR2

  • Sequences in target group in PR2: 4633

  • Sequences in target group in PR2 that need update: 4844

  • Updated sequences that are not active in PR2: 32 (sequences removed from PR2)

  • Sequences from target group that are not in the update: 9 (these sequences have been removed from PR2 because too many ambiguities)

  • Sequences duplicated: 0

  • chimera sequences: 343

  filter(pr2_18S, !!as.symbol(target_level) %in% target_group) %>% 
     DT::datatable(caption = "Sequences of target group in pr2 ") 

  filter(pr2_18S, (genbank_accession %in% pr2_update$genbank_accession)) %>% 
    DT::datatable(caption = "Sequences of PR2 that need update")
  
  filter(pr2_update, (genbank_accession %in% pr2_18S_removed$genbank_accession)) %>% 
     DT::datatable(caption = "Updated sequences that are not active in PR2")
    
  filter(pr2_update, !(genbank_accession %in% pr2_18S$genbank_accession))%>% 
     DT::datatable(caption = "Updated sequences that are not present in PR2 ")
  
  filter(pr2_18S, (!!as.symbol(target_level) %in% target_group) & 
             !(genbank_accession %in% pr2_update$genbank_accession)) %>% 
  DT::datatable(caption = "Sequences from target group in PR2 that are not in update")
  
  left_join(select(pr2_update, genbank_accession), 
            select(pr2_18S, genbank_accession, pr2_accession)) %>% 
   count(genbank_accession) %>% 
   filter(n > 1) %>% 
   DT::datatable(caption = "Sequences updated with 2 entries in PR2  (e.g. with and without introns) ")
  
   filter(pr2_update, !is.na(chimera)) %>% 
     DT::datatable(caption = "Chimera")

5 Taxonomy

5.1 Build and check

  • The old taxonomy is removed and the new taxonomy replaces it…
pr2_taxo_updated <- pr2_update %>% 
  filter(is.na(chimera)) %>% 
  group_by_at(pr2.env$taxo_levels[[8]][4:8]) %>%  
  count() 

pr2_taxo_check(pr2_taxo_updated, pr2.env$taxo_levels[[8]][4:8], full_path("taxo"))

pr2_taxo_updated <- pr2_taxo_updated %>% 
  rename_all(~ str_c(.,"_new" )) %>%
  dplyr::rename(species = species_new) %>% 
  left_join(pr2_taxo) %>% 
  rename_at(pr2.env$taxo_levels[[8]], ~ str_c(.,"_old" )) %>% 
  rename_all( ~ str_replace(.,"_new", "_8" )) %>% 
  dplyr::rename(species_8 = species_old) %>% 
  mutate(kingdom_8 = "Eukaryota",
         supergroup_8 = "Rhizaria",
         division_8 = "Radiolaria",
         domain_9 = "Eukaryota",
         supergroup_9 = "TSAR",
         division_9 = "Rhizaria",
         subdivision_9 = "Radiolaria",
         class_9 = class_8,
         order_9 = order_8,
         family_9 = family_8, 
         genus_9 = genus_8,
         species_9 = species_8,
         taxo_edited_version = pr2.env$version,
         taxo_edited_by = "M. Sendin")  
  
export(pr2_taxo_updated, full_path("taxo/taxo_updated.xlsx"))

5.2 Find taxa in PR2 that are not included in the update - DO NOT APPLY

  • The old taxonomy is removed
pr2_taxo_not_updated <- pr2_taxo %>% 
  filter(!!as.symbol(target_level) %in% target_group) %>% 
  filter(!(species %in% pr2_taxo_updated$species_8))

export(pr2_taxo_not_updated , full_path("taxo_not_updated.xlsx"))

6 Finalization

6.1 Sequences that need updating

pr2_update_final <- pr2_update %>% 
  select(genbank_accession, species, start, end, chimera, eukref_biotic_relationship) %>% 
  dplyr::rename(species_new = species, 
                start_new = start,
                end_new = end) %>% 
  left_join(select(pr2_main, pr2_accession, genbank_accession, species, start, end, removed_version, edited_remark))

6.2 Sequences with different start or end

  • 9 sequences have been reloaded into PR2 using the Genbank script.
  • These sequences had been initially uploaded with Silva which had imperfect trimming of the end.
pr2_sequence_to_update <- pr2_update_final %>% 
  filter((start != start_new)|(end != end_new)) 

pr2_sequence_to_update %>% 
  DT::datatable(caption = "Actual sequence and pr2_accession needs to be updated  ")

6.3 pr2_main: Sequences without species name or with different species

  • Number of updated sequences with prior taxonomy: 4399
  • Number of updated sequences without prior taxonomy 220
pr2_main_updated <- pr2_update_final %>% 
  filter((species != species_new)|is.na(species) |!is.na(chimera)) 

pr2_main_updated %>% 
  DT::datatable(caption = "Actual sequence and pr2_accession needs to be updated  ")

glue::glue("Number of updated sequences  with prior taxonomy: {nrow(filter(pr2_main_updated, !is.na(species)))}")
glue::glue("Number of updated sequences  without prior taxonomy {nrow(filter(pr2_main_updated, is.na(species)))}")

pr2_main_updated <- pr2_main_updated %>%  
    select (genbank_accession, 
            species_8 = species_new, 
            species_9 = species_new, 
            removed_version,
            edited_remark,
            chimera) %>% 
    mutate(edited_version = pr2.env$version, 
           edited_by = pr2.env$editor,
           chimera = case_when(!is.na(chimera) ~ 1),
           chimera_remark = case_when(!is.na(chimera) ~ "chimera detected by M. Sandin"),
           removed_version = case_when(!is.na(chimera) ~ pr2.env$version,
                                       TRUE ~ removed_version ))

6.3.1 pr2_metaddata

pr2_metadata_updated <- pr2_update_final %>%
  select(genbank_accession, eukref_biotic_relationship) %>% 
  filter(!is.na(eukref_biotic_relationship))

7 Save everything to an Excel file

 file_pr2_imports <-  full_path("pr2_update_tables.xlsx")
 onglets <- list("pr2_main_updated" = pr2_main_updated,
                 "pr2_metadata_updated" = pr2_metadata_updated,
                 "pr2_taxo_updated" = pr2_taxo_updated
                 )
 export(onglets, file_pr2_imports)