Skip to content

spficklin/BDSS

 
 

Repository files navigation

What is the Big Data Smart Socket (BDSS)?

The increasing size of datasets used in scientific computing has made it difficult or impossible for a researcher to store all their data at the compute site they are using to process it. This has necessitated that a data transfer step become a key consideration in experimental design. Accordingly, scientific data repositories such as NCBI have begun to offer services such as dedicated data transfer machines and advanced transfer clients. Despite this, many researchers continue familiar but suboptimal practices: using slow transfer clients like a web browser or scp, transferring data over wireless networks, etc.

BDSS aims to alleviate this problem by shifting the burden of learning about alternative file mirrors, transfer clients, tuning parameters, etc. from the end user researcher to a group of "data curators". It consists of three parts:

Components

  • Metadata repository
    • Central database managed by data curators
    • Matches patterns of data file URLs and maps them to alternate sources
    • Includes information about the transfer tool to use to retrieve the data
  • BDSS transfer client
    • Consumes information from metadata repository
    • Invokes transfer tools
    • Reports analytics to metadata repository
  • Integration as a Galaxy data transfer tool

Get Started

Examples

All examples here require a metadata repository configured to support them. The default metadata repository at https://bdss.bioinfo.wsu.edu/ supports these examples and the necessary configuration is also listed here.

NCBI SRA archive

NCBI makes files available for transfer using Aspera Connect, a tool with "improved data transfer characteristics" vs FTP or HTTP. If ascp is installed on your machine, BDSS can handle building the appropriate command.

Without BDSS:

ascp -i $HOME/.aspera/connect/etc/asperaweb_id_dsa.openssh -T [email protected]:/sra/sra-instant/reads/ByRun/sra/SRR/SRR039/SRR039885/SRR039885.sra ./

With BDSS:

bdss transfer -u 'ftp://ftp.ncbi.nlm.nih.gov/sra/sra-instant/reads/ByRun/sra/SRR/SRR039/SRR039885/SRR039885.sra'

Metadata repository configuration:

{
  "data_sources": [
    {
      "description": "",
      "label": "NCBI Sequence Read Archive with FTP",
      "test_files": [],
      "transfer_mechanism": {
        "options": {},
        "type": "curl"
      },
      "transforms": [
        {
          "for_destinations": [],
          "options": {
            "new_scheme": "aspera"
          },
          "target": "NCBI Sequence Read Archive with Aspera",
          "type": "change_scheme"
        }
      ],
      "url_matchers": [
        {
          "options": {
            "pattern": "^ftp://ftp\\.ncbi\\.nlm\\.nih\\.gov/sra"
          },
          "type": "regular_expression"
        }
      ]
    },
    {
      "description": "",
      "label": "NCBI Sequence Read Archive with Aspera",
      "test_files": [],
      "transfer_mechanism": {
        "options": {
          "disable_encryption": true,
          "username": "anonftp"
        },
        "type": "aspera"
      },
      "transforms": [],
      "url_matchers": [
        {
          "options": {
            "pattern": "^aspera://ftp\\.ncbi\\.nlm\\.nih\\.gov/sra"
          },
          "type": "regular_expression"
        }
      ]
    }
  ],
  "destinations": []
}

JGI Genome Portal

To download files from the JGI Genome Portal, you must first authenticate. BDSS can prompt for credentials and handle storing your session cookies.

Without BDSS:

curl 'https://signon.jgi.doe.gov/signon/create' --data-urlencode 'login=USER_NAME' --data-urlencode 'password=USER_PASSWORD' -c cookies > /dev/null
curl 'http://genome.jgi.doe.gov/ext-api/downloads/get-directory?organism=PhytozomeV10' -b cookies > get-directory

With BDSS:

bdss transfer -u 'http://genome.jgi.doe.gov/ext-api/downloads/get-directory?organism=PhytozomeV10'
JGI Genome Portal username?USER_NAME
JGI Genome Portal password?USER_PASSWORD

Metadata repository configuration:

{
  "data_sources": [
    {
      "description": "",
      "label": "JGI Genome Portal",
      "test_files": [],
      "transfer_mechanism": {
        "options": {
          "auth_url": "https://signon.jgi.doe.gov/signon/create",
          "password_field": "password",
          "password_prompt": "JGI Genome Portal password?",
          "username_field": "login",
          "username_prompt": "JGI Genome Portal username?"
        },
        "type": "session_authenticated_curl"
      },
      "transforms": [],
      "url_matchers": [
        {
          "options": {
            "pattern": "http:\\/\\/genome\\.jgi\\.doe\\.gov\\/ext-api"
          },
          "type": "regular_expression"
        }
      ]
    }
  ],
  "destinations": []
}

About

Big Data Smart Socket

Resources

License

Stars

Watchers

Forks

Packages

No packages published

Languages

  • Python 83.7%
  • HTML 13.3%
  • Shell 2.8%
  • Other 0.2%