Skip to content

ukhsa-collaboration/ena_submission

Repository files navigation

README for ena_submission tool


This tool in a single step will create the necessary xml files and uploads all your files to the ENA repository.

Description of the Process


In order to submit your files as a batch to ENA, you need to create some files that have information about your samples which then you can use to upload to ENA. It can be a tedious task. This script would make this possible in one step. This is how it works (provided that you have provided the necessary files, see 'Required files' below):

  1. Creates a checksum file of all your files needed for submission.
  2. Uploads the files to the ENA server using FTP. It will use the authentications you have provided it.
  3. Provided the data format is correct, the script will create five xml files: sample.xml, experiment.xml, run.xml, study.xml and submission.xml.
  4. If all the .xml files have been successfully created, the script will run a test curl command that will upload all the .xml files to the ENA server and returns a receipt.xml file. If this process failes, you will be notified and you must correct the error and re-submit using the -curl option. The script will run a test command to test if its sucessful. You must then run the production curl command to upload your data to production.

Dependencies


Python >= 2.7

Required files


To upload your files to ENA, you will need the following:

  1. A directory which contains all your files. If you are uploading fastq files, the files must be in the following format (see examples in the examples folder):

SAMPLE_NAME.R1.fastq.gz

SAMPLE_NAME.R2.fastq.gz

E.g.

MN127.R1.fastq.gz

MN127.R2.fastq.gz

Otherwise, if you are uploading a different type of file, like bam files or fasta files, you can name them whatever you like but they should have the prefix at the end, e.g. .bam or .fasta. Use the -F flag to do so, e.g. -F bam.

  1. A csv (comma separated values) file that contains all the information for each of your samples. The text file should contain four required columns. You may add as many columns of data as you wish after the fourth required column. The columns are separated by commas. e.g. (note the headings are required as seen below):

SAMPLE,TAXON_ID,SCIENTIFIC_NAME,DESCRIPTION

data....

You may like to use the template file provided in the examples folder. You might also like to see an example file in the examples folder. Please note that ENA now check that the Scientific_name should be matching to taxon_id so please make sure you write the correct scientific_name.

  1. A text file that contains the title for your project and an abstract explaining the project. The two columns must be separated by tabs. There are no headings needed here. e.g (see examples in the examples folder):

Bioinformatics for biologists \t Bioinformatics is great for biologists.....etc

  1. An ENA ftp username and password. If you do not have one, you need to contact ENA at: [email protected].

NOTE!! If you are uploading files other than fastq files, you must use the -F flag and add the prefix, e.g. -F bam.

Before you run the script....


Before you run the script, make sure you have the following ready as you cannot proceed without it:

-i : full path to your files you would like to upload to ENA. (point 1 above)
-r : a reference name for your project
-f : the path to the data file (point 2 above)
-a : the path to the title and abstract text file (point 3 above)
-o : the path to the output directory
-user : ENA username (point 4 above)
-pass : ENA password (point 4 above)

You should familarise yourself with the other options before you begin running the script. To do so, run python ena_submission.py -h

Examples of commands


If you like to run a single command that does EVERYTHING, i.e. creates checksums, uploads your files to ftp, creates the xml files, uploads the xml files and run the curl command, then use the -d option, e.g. |n

E.g. for fastq submission (otherwise use -F suffix)|n

  python ena_submission.py -d -i /PATH/TO/SAMPLES_FILES_DIR -r phe_mycoplasma -f /PATH/TO/DATA_FILE.txt -o /PATH/TO/OUTPUT_DIR -a /PATH/TO/TITLE_ABSTRACT_FILE.txt -c CENTRE_NAME -user Webin-YYYYY -pass XXXXX 

You may wish to run each step individually. These are the different steps you may run manually:


To just create the checksums file:

--create_checksums_file : only create the checksums file for my fastqs. e.g.

python ena_submission.py -cs -i /PATH/TO/SAMPLES_FILES_DIR -r test3

To just upload your data to the ENA server:

--upload_data_to_ena_ftp_server : only upload the data to ENA. e.g.

python ena_submission.py -ftp -i /PATH/TO/SAMPLES_FILES_DIR -r NAME_OF_YOUR_PROJECT -user Webin-YYYYY -pass XXXXX

To just create sample.xml file:

-x sample : use the word 'sample' (same goes for the rest) to create sample.xml only. e.g.

python ena_submission.py -x sample -i /PATH/TO/SAMPLES_FILES_DIR -r NAME_OF_YOUR_PROJECT -f /PATH/TO/DATA_FILE.txt -c YOUR_CENTRE_NAME -o /PATH/TO/OUTPUT_DIR

To just create experiment.xml file:

-x experiment : to create experiment.xml only. e.g.

python ena_submission.py -x experiment -i /PATH/TO/SAMPLES_FILES_DIR -r NAME_OF_YOUR_PROJECT -f /PATH/TO/DATA_FILE.txt -c YOUR_CENTRE_NAME -o /PATH/TO/OUTPUT_DIR  

To just create run.xml file:

-x run : to create run.xml only. e.g.

python ena_submission.py -x run -i /PATH/TO/SAMPLES_FILES_DIR -r NAME_OF_YOUR_PROJECT -c YOUR_CENTRE_NAME -o /PATH/TO/OUTPUT_DIR

To just create study.xml file: -x study : to create study.xml only. e.g.

python ena_submission.py -x study -a /PATH/TO/TITLE_ABSTRACT_FILE.txt -r NAME_OF_YOUR_PROJECT -i /PATH/TO/SAMPLES_FILES_DIR -c YOUR_CENTRE_NAME -o /PATH/TO/OUTPUT_DIR

To just create submission.xml file:

-x submission : to create submission.xml only. e.g.

python ena_submission.py -x submission -i /PATH/TO/SAMPLES_FILES_DIR -o /PATH/TO/OUTPUT_DIR -r NAME_OF_YOUR_PROJECT -c YOUR_CENTRE_NAME

NOTE: Here you may hold your data from release publicly if you use -ho option and specify the date.

To just run the curl command:

-curl : just run curl command. Note: you must have uploaded the data and created all the xml files to run the curl command. e.g.

python ena_submission.py -curl -user Webin-YYYYY -pass XXXX -o /PATH/TO/OUTPUT_DIR

Usage



-h, --help            show this help message and exit
--generate_xml_file_for GENERATE_XML_FILE_FOR, -x GENERATE_XML_FILE_FOR
                      please provide the name for the xml you like to
                      generate, one of the
                      following:all,experiment,run,sample,study,submission
--dir_of_input_data DIR_OF_INPUT_DATA, -i DIR_OF_INPUT_DATA
                      please provide the path for the files you like to
                      upload to ENA. NOTE: the fastq files must be in the
                      following format: *.R1.fastq.gz and *.R2.fastq.gz.
--data_file DATA_FILE, -f DATA_FILE
                      data_file, file : this text file must have at least
                      four columns seperated by TABS in the following order
                      and with the following headings: Column 1: SAMPLE,
                      Column 2: TAXON_ID, Column 3: SCIENTIFIC_NAME, Column
                      4: DESCRIPTION. If you like to add further data then
                      add it after the DESCRIPTION column.
--ftp_user_name FTP_USER_NAME, -user FTP_USER_NAME
                      please provide the ftp user name
--ftp_password FTP_PASSWORD, -pass FTP_PASSWORD
                      please provide the ftp password
--title_and_abstract_file TITLE_AND_ABSTRACT_FILE, -a TITLE_AND_ABSTRACT_FILE
                      please provide the title and abstract text file. This
                      is needed to generate the study.xml file. The file
                      should be in the following format: full title of the
                      project abstract
--center_name CENTER_NAME, -c CENTER_NAME
                      Please provide the center name
--refname REFNAME, -r REFNAME
                      Please provide the unique name for the whole
                      submission. This name must not have been used before
                      in any other submission to ENA.
--library_strategy LIBRARY_STRATEGY, -s LIBRARY_STRATEGY
                      please provide the library strategy used. default =
                      WGS
--library_source LIBRARY_SOURCE, -u LIBRARY_SOURCE
                      please provide the library source used. default =
                      GENOMIC
--library_selection LIBRARY_SELECTION, -e LIBRARY_SELECTION
                      please provide the library selection used. default =
                      RANDOM
--read_length READ_LENGTH, -l READ_LENGTH
                      Please provide the read length. default = 100
--read_sdev READ_SDEV, -sdev READ_SDEV
                      Please provide the read stdv. default = 0.0
--instrument_model INSTRUMENT_MODEL, -m INSTRUMENT_MODEL
                      please provide the name of the instrument model used.
                      default = Illumina HiSeq 2500
--filetype FILETYPE, -F FILETYPE
                      Please provide the file type, default=fastq
--hold_date HOLD_DATE, -ho HOLD_DATE
                      This option will hold the data privately until a
                      specified date. Please provide the specified date
                      using the following format: YYYY-MM-DD
--release, -R         This option will release the data immediatley to the
                      public domain. Cannot be used with --hold_date.
                      default = False
--curl_command, -curl
                      Run curl command to upload everything to ENA
--create_checksums_file, -cs
                      Create a checksum file
--upload_data_to_ena_ftp_server, -ftp
                      Upload data into the ftp server
--out_dir OUT_DIR, -o OUT_DIR
                      please provide the path to the output directory which
                      will include all the xml files.

Contact

Ali Al-Shahib [email protected]

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages