Methods in Marine Biodiversity

This is a collection of files for the Marine Biodiversity Methods course at Friday Habor Labs.

Methods

Taxonomic Annotation

We applied a number of quality filters to the sequence data (everything up to pre-clustering); now the aim is to apply a taxon name to each sequence.

Our approach for taxonomic annotation is the following:

Does the sequence have a match in the FHL course barcode database at >97% identity? (blastn)
- YES:
  - Is there only one hit at that level?
    - YES: use the name associated with that specimen
    - NO: use multiple names, resolving differences by (???...???)
- NO:
  - Does the sequence match a sequence in GenBank at >97% identity? (blastn)
    - YES:
      - Is there only one hit at that level?
        
        YES: use the name associated with that specimen
        
        NO: use multiple names, resolving differences by (???...???)
    - NO:
      - After clustering at 5% (CROP), does OTU sequence have a match in GenBank at >85% identity? (blastn)
        
        YES: use the phylum name associated with that GenBank sequence
        
        NO:
        
        Does the sequence have a probability of assignment to a sequence in (database?) with posterior probability >0.8?
        
        YES: use the phylum name associated with that taxon
        
        NO: sequence is labeled as 'unidentified taxon'

The output of this process is a table with the following columns; one row for each sequence in the pre-clustering fasta file:

seq_id [char]: sequence header from pre-clustering fasta file
FHL_db [boolean]: hit in local database?
GB_95 [boolean]: hit in GenBank >95%id?
CROP_ID [char]: sequence header of the OTU fasta file
GB_85 [boolean]: hit in GenBank >85%id after clustering?
SAP_prob [numeric,{0,1}]: posterior probability of assignment to a taxon using SAP
taxon_name [char]: taxon name