Skip to content

Latest commit

 

History

History
71 lines (54 loc) · 3.27 KB

README.rst

File metadata and controls

71 lines (54 loc) · 3.27 KB
Raw

Default ITS1 database

We include a default ITS1 database with THAPBI PICT using the following four sets of biological sequences (subject to taxonomy filtering):

  • Curated Phytophthora ITS1 sequences (mostly at species level) from file database/Phytophthora_ITS1_curated.fasta. This is sorted by sequence. Examples sharing the same trimmed sequence are recorded as a single record with multiple semi-colon separated accession genus species entries. Sadly many published sequences are missing the first 32bp which our default primers would amplify, namely TTTCCGTAGGTGAACCTGCGGAAGGATCATTA. This has been assumed and added to the FASTA in lower case as needed - at the time of writing exceptions are rare.
  • Curated Nothophytophthora ITS1 sequences (mostly at species level) from file Nothophytophthora_ITS1_curated.fasta. Again, most have been extended with the assumed 32bp leader.
  • Curated other Peronosporales ITS1 sequences including Peronosclerospora from file Peronosporales_ITS1_curated.fasta, currently imported only to genus. Most have been extended with the assumed 32bp leader, or in some cases the final dozen bases.

  • NCBI Peronosporales (including Phytophthora) or Pythiales at genus level, in two files extracted from tens of thousands of candidate sequences in file Oomycota_ITS1_search.fasta (not under version control) downloaded from an NCBI Entrez search via script Oomycota_ITS1_search.sh:

    (Peronosporales[organism] OR Pythiales[organism])
AND ((internal AND transcribed AND spacer) OR its1)
AND 150:10000[sequence length]

    The search results are filtered to extract the amplicons with the expected 32bp leader (allowing for the leading T or TT to be missing) with cutadapt as file Oomycota_ITS1_w32.fasta (under version control).

    Then script ../scripts/missed_refs.py is used to catch useful references without the typical 32bp leader in full as file Oomycota_ITS1_obs.fasta (under version control), where observed in at least five of our samples (using file unknowns.fasta which is not under version controls but is generated from our sequencing data using script ../scripts/unknowns.py).

    Note the import command discards uncultured entries.

  • Observed ITS1 sequence files single_isolates/*.fasta with at least abundance 1000 (well above plate negative controls) from single isolate positive controls run on MiSeq plates.

Additionally:

  • Four G-BLOCKS synthetic controls in file database/controls.fasta
  • This used the NCBI taxonomy recorded in the script, which typically means a handful of unrecognised curated entries may be imported without an NCBI taxid.

The database is created with the database/build_ITS1_DB.sh script:

$ cd database/
$ ./build_ITS1_DB.sh
$ chmod a-w ITS1_DB.sqlite
$ cp ITS1_DB.sqlite ../thapbi_pict/ITS1_DB.sqlite

We include the binary SQLite3 file as thapbi_pict/ITS1_DB.sqlite in the software releases, but that binary SQLite3 database file itself is not under version control. Instead, we track a plain text FASTA dump of the database as database/ITS1_DB.fasta giving a meaningful change history (we initially tracked this as a plain text SQL dump, database/ITS1_DB.sql).