Skip to content

Latest commit

 

History

History
131 lines (123 loc) · 5.53 KB

cookbook.md

File metadata and controls

131 lines (123 loc) · 5.53 KB
Raw

Cookbook

Custom reference database

From a DFAST reference file

  1. Create a reference file
    Prepare a TAB-separated reference file following the format described in the Reference Database section.
    The file extension must be '.ref', say 'your_reference.ref'.

  2. Format database

    python $DFAST_APP_ROOT/scripts/reference_util.py formatdb -i your_reference.ref

    Then, index files for GHOSTX and BLASTP will be generated in the same location as the reference file.

  3. Configure and run
    Specify the reference file in the database attribute in the 'DBsearch' section,
    e.g. "database": "/path/to/your_reference.ref"

    Alternatively, you can run dfast using the --database option.

    dfast --genome your_genome.fa --database /path/to/your_reference.ref

From a FASTA file

You can prepare a database easily from a FASTA file using 'reference_util.py'.
The script can parse FASTA definition lines for NCBI/UniprotKB/Prokka styles.

  1. Convert a FASTA file into DFAST reference format 
    python $DFAST_APP_ROOT/scripts/reference_util.py fasta2dfast -i your_reference.fasta -o your_reference.ref
  2. Format database, configure, and run
    Then, follow the same procedure as above.

OrthoSearch

OrthoSearch identifies orthologous genes based on a simple Reciprocal-Best-Hit (RBH) approach. This is effective in reducing running time and in transferring annotations from a reference genome of the closely-related organism.
This recipe shows how to perform OrthoSearch.

Reference proteome

OrthoSearch requires a 'reference proteome' file that contains all protein sequences in a genome. The file format must be either of FASTA, GenBank, or DFAST reference format. In addition to a plain FASTA format (sequence ID and definition), OrthoSearch can parse FASTA definition lines of UniProt, GenBank, and Prokka styles.
The format is automatically recognized.

Recipe

Our recommendation is to download a GenBank-format file from the NCBI Assembly Database and to use it as a reference.

  1. Download a reference proteome 
    python $DFAST_APP_ROOT/scripts/file_downloader.py --assembly GCF_000005845
    This will download the latest version of the Escherichia coli str. K-12 genome in a GenBank-format into the current directory with the file name 'GCF_000005845.2.gbk'. You can use the --out option to specify the directory into which the file is downloaded.
  2. Run DFAST
    Use --references to specify the reference proteome(s). 
    dfast --genome your_genome.fa --references GCF_000005845.2.gbk
    You can specify multiple proteome files with commas to separate files. 
    dfast --genome your_genome.fa --references GCF_000005845.2.gbk,GCA_000008865.1.faa
    When multiple files are used as references, all-vs-all alignments are conducted between a query proteome and each of the reference proteomes, and the highest-scoring hit will be adopted as the result.
  • Configuration
    Reference proteomes can also be specified in the configuration file.
    Set enabled to True, and specify references in the 'FUNCTIONAL_ANNOTATION' part. 
    {
    "component_name": "OrthoSearch",
    "enabled": True,
    "options": {
        # "cpu": 2,  # Uncomment this to set the component-specific number of CPUs.
        "skipAnnotatedFeatures": False,
        "evalue_cutoff": 1e-6,
        "qcov_cutoff": 75,
        "scov_cutoff": 75,
        "aligner": "ghostx",
        "aligner_options": {},
        "references": ["GCF_000005845.2.gbk", "GCA_000008865.1.faa"]
    },
},

BlastSearch

BlastSearch is for protein homology search against a large-sized reference database, such as pre-formatted Blast databases like RefSeq Protein and SwissProt available at the NCBI FTP site.

How to use a pre-formatted BLAST database

  1. Download a database from NCBI 
    wget ftp://ftp.ncbi.nlm.nih.gov//blast/db/swissprot.tar.gz
tar xvfz swissprot.tar.gz
  2. Create a configuration file
    Set enabled to True, and specify database to be searched against. You can also specify dbtype, but normally, leaving it 'auto' will do.
    Place this part upstream of 'DBsearch' against the default database if you want to give priority to 'BlastSearch'. 
    {
    "component_name": "BlastSearch",
    "enabled": True,
    "options": {
        # "cpu": 2,  # Uncomment this to set the component-specific number of CPUs.
        "skipAnnotatedFeatures": False,
        "evalue_cutoff": 1e-6,
        "qcov_cutoff": 75,
        "scov_cutoff": 75,
        "aligner": "blastp",  # Must be blastp
        "aligner_options": {},
        "dbtype": "auto",  # Must be either of auto/ncbi/uniprot/plain
        "database": "/path/to/swissprot",
    },
},
  3. Run DFAST 
    dfast --genome your_genome.fa --config your_config.py

Prepare database from a FASTA file

Here is an example to create a database for RefSeq nonredundant archaeal proteins.

  1. Download FASTA files 
    wget ftp://ftp.ncbi.nlm.nih.gov//refseq/release/archaea/archaea.nonredundant_protein.*.protein.faa.gz
gunzip -c archaea.nonredundant_protein.*.protein.faa.gz > archaea.nonredundant_protein.faa
  2. Format database
    Be sure to use -parse_seqids. 
    makeblastdb -hash_index -parse_seqids -dbtype prot -in archaea.nonredundant_protein.faa
  3. Create a configuration file and run DFAST
    Follow the recipe described above.