Medline Ranker REST API beta!

REST API

Please read carefully the documentation, citation and disclaimer before using this service. In order to query the web service, an HTTP request has to be sent to a specific URL. This URL is composed by a base URL followed by key-value pairs, the latter defining the input parameters of the queried service. The resulting file is a table of genes or diseases as tab-separated values. Please use this service moderately, letting reasonable time between consecutive queries. Upon heavy or unexpected use of the service, its access may be blocked to specific IP ranges without notice. Please note that this service is a non stable beta version. Especially, it does not produce the full output as expected on the DTD (missing parameters information).

Base URL

  • http://cbdm-01.zdv.uni-mainz.de/~jfontain/cgi-bin/ranker.pl

Template full URL

Please note the question mark (?) to separate the Base URL to the rest, symbols & to separate key-value pairs, and symbols = to separate keys and values.
  • BaseURL?key1=value1&key2=value2&key3=value3 ...

Parameters and corresponding key-value pairs

Please see the documentation for a definition of each parameter: Documentation page.
  • Webservice
    • key: webservice
    • possible values:
      • 1
  • Training set type
    • key: trset
    • possible values:
      • pubmed: then key trset_text must be a standard free text query as in PubMed
      • allmesh: then key trset_text must be a semi-colon separated list of MeSH terms
      • list: then key trset_text must be a semi-colon separated list of PMIDs
  • Training set value
    • key: trset_text
    • possible values:
      • if trset=pubmed: a standard free text query as in PubMed
      • if trset=allmesh: a semi-colon separated list of exact MeSH terms
      • if trset=list: a semi-colon separated list of PMIDs
  • Test set type
    • key: tset
    • possible values:
      • list: abstracts provided by the user; then key tset_text must be a semi-colon separated list of PMIDs
      • trset: abstracts from the training set; the key trset_text must be empty
      • bgset: abstracts from the background set; the key trset_text must be empty
      • last_months: abstracts published during the last N months; the value of key trset_text must be set to N
      • last_years: abstracts published during the last N years; the value of key trset_text must be set to N
  • Test set value
    • key: tset_text
    • possible values:
      • if tset=list: a semi-colon separated list of PMIDs
      • if tset=trset: must be empty
      • if tset=bgset: must be empty
      • if tset=last_months: an integer value for a number of months
      • if tset=last_years: an integer value for a number of years
  • Background set type
    • key: bgset
    • possible values:
      • list: abstracts provided by the user; then key bgset_text must be a semi-colon separated list of PMIDs
      • medline: abstracts from the all Medline/PubMed; the key bgset_text must be empty
  • Background set value
    • key: bgset_text
    • possible values:
      • if bgset=list: a semi-colon separated list of PMIDs
      • if medline=trset: must be empty
  • Target database
    • key: target_db
    • possible values:
      • medline: uses Medline/PubMed abstracts in the test set
      • pmc: limits the test set to PubMed Central abstracts
  • Minimum word weight
    • key: weight
    • possible values:
      • positive float value (e.g. 1.5 or 3.75)
  • Output

    The output of this service is XML data defined in a DTD document at: http://cbdm-01.zdv.uni-mainz.de/~jfontain/soap/dtd/medlineranker_beta2.dtd

    Examples

    The key-value pairs below are required for the query "Search abstracts related to Alzheimer disease in the last 3 months of PubMed indexed abstracts"
    • webservice=1
    • trset=pubmed
    • trset_text=Alzheimer disease
    • tset=last_months
    • tset_text=3
    Then the full URL to query is the following: An other example: rank 1 year of PubMed for the same topic:

Comments are closed.