DESCRIPTION:

 Scripts for automatic download genbank records based on file with list of accessionn numbers (or directory with those files).
 The files have to have '.lst' extension. Furthermore it allows to dived the genbank files into genes and genome sequences
 and taxonomy table adapted for blast or qiime2 classifier analysis.

  DATE:      01.07.2025
  VERSION:  1.0
  CREATOR:  T.J.SANKO
  ADDRESS:  Stellenbosch University, Tygerberg Campus,
            Francie van Zijl Drive, Cape Town, South Africa
  CONTACT:  tjsanko@sun.ac.za

1. Requirements

  Script is design to work on any Linux machine with bash/shell language.
  It was tested on CentOS7 and Ubuntu SMP 24.04.2.

2. Installation

  Uses basic bash/shell commands. Does not require any additional installation.
  May require changing permissions after downloading to linux/unix environment
 

chmod +x *.sh

3. Syntax

./acc2db.sh -i file.lst -D -C
 or alternative:

 ./acc2db.sh --dir -D -C

4. Options

  Required:
    -i | --infile       input file with accession number as a list. Has to have
                       '.lst' extension. Overrides '--dir' option.

    -d | --dir          Alternative to '-i' option. Does not require any
                        parameters. Searches through current directory.
                        Inferrior to '-i' and cannot be used together with that
                        option.
  Optional:
    -D | --download     Automatically downloads the files, else it creates
                        script that can be use to download manually
                        (default: OFF)

    -C | --clean        Automatically remove all intermediate file, else it
                        preserves them (default: OFF)

    -v | --version      prints current version

    -h | --help         prints this message

5. Description

  acc2db script takes as input list of accession numbers. The list has to be in file and has to have '.lst' extension.
  Alternatively, instead a file, current directory can be provided by passing the '-d' parameter. It does not search
  recursively (limited depth to one level only). If both, '-d' and '-i' are provided, '-i' takes priority. The '-d'
  parameter allows for more than one list to be used as long as it is in the current directory. After all lists are read,
  all duplicated accessions are removed and only unique numbers are divided into download lines and saved in
  'download_TTTTNN.sh' files. Each file has max 300 accession numbers per line, max 500 lines per file. TTTT in the name
  is the file format to download (e.g. gb, gff, asn1, json) and NN is the consecutive number. If '-D' parameter is not
  provided the files will be created with instructions inside on hot to run them manually. Last line in each file checks
  if 'check_download' script exists in current location, and pass list of accessions, from all lines in the file. It will
  be used for redonload of the record if it was not successful in the first round. Each 'download_TTTTNN.sh' file will
  pull records per all lines and will save them under 'current_dir/ncbi_sl' directory.  Successful download and
  (eventual) download of failed records will be mereged into one file 'ncbi_DDD.TTTT' where DDD is the name of the
  database and TTTT is format of files downloaded. If '-C' option is provided it will clean up after whole process
  removing the 'download_TTTTNN.sh' files and 'current_dir/ncbi_dl' folder leaving only the final file.

6. Personalization

 Some of the defaults can be edited to suit better the needs of the user.
 The defaults that can be adapted are:

PER_FILE	Divides files in N-download lines per file (default 500)
PER_LINE	No. of accession numbers downloaded at once per line (default 300). Do not exceed over 300 else it will block the connection to NCBI and you will end with many failed downloads
DB	NCBI database to download from: nuccore, protein, pubmed (default 'nuccore')
TYPE	database format: gb, gff, asn1, json (default 'gb')
MODE	database mode: text, summary (default 'text')
OUT	output directory name (default 'current_dir/ncbi_dl')