bio 0.7.0

data/lib/bio/io/pubmed.rb

@@ -0,0 +1,189 @@
+#
+# bio/io/pubmed.rb - NCBI Entrez/PubMed client module
+#
+#   Copyright (C) 2001 KATAYAMA Toshiaki <k@bioruby.org>
+#
+#  This library is free software; you can redistribute it and/or
+#  modify it under the terms of the GNU Lesser General Public
+#  License as published by the Free Software Foundation; either
+#  version 2 of the License, or (at your option) any later version.
+#
+#  This library is distributed in the hope that it will be useful,
+#  but WITHOUT ANY WARRANTY; without even the implied warranty of
+#  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+#  Lesser General Public License for more details.
+#
+#  You should have received a copy of the GNU Lesser General Public
+#  License along with this library; if not, write to the Free Software
+#  Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307  USA
+#
+#  $Id: pubmed.rb,v 1.12 2005/09/08 01:22:12 k Exp $
+#
+
+require 'net/http'
+require 'cgi' unless defined?(CGI)
+
+module Bio
+
+  class PubMed
+
+    def self.query(id)
+      host = "www.ncbi.nlm.nih.gov"
+      path = "/entrez/query.fcgi?tool=bioruby&cmd=Text&dopt=MEDLINE&db=PubMed&uid="
+
+      http = Net::HTTP.new(host)
+      response, = http.get(path + id.to_s)
+      result = response.body
+      if result =~ /#{id}\s+Error/
+        raise( result )
+      else
+        result = result.gsub("\r", "\n").squeeze("\n").gsub(/<\/?pre>/, '')
+        return result
+      end
+    end
+
+    def self.pmfetch(id)
+      host = "www.ncbi.nlm.nih.gov"
+      path = "/entrez/utils/pmfetch.fcgi?tool=bioruby&mode=text&report=medline&db=PubMed&id="
+
+      http = Net::HTTP.new(host)
+      response, = http.get(path + id.to_s)
+      result = response.body
+      if result =~ /#{id}\s+Error/
+        raise( result )
+      else
+        result = result.gsub("\r", "\n").squeeze("\n").gsub(/<\/?pre>/, '')
+        return result
+      end
+    end
+
+    def self.search(str)
+      host = "www.ncbi.nlm.nih.gov"
+      path = "/entrez/query.fcgi?tool=bioruby&cmd=Search&doptcmdl=MEDLINE&db=PubMed&term="
+
+      http = Net::HTTP.new(host)
+      response, = http.get(path + CGI.escape(str))
+      result = response.body
+      result = result.gsub("\r", "\n").squeeze("\n")
+      result = result.scan(/<pre>(.*?)<\/pre>/m).flatten
+      return result
+    end
+
+    def self.esearch(str, hash = {})
+      hash['retmax'] = 100 unless hash['retmax']
+
+      opts = []
+      hash.each do |k, v|
+        opts << "#{k}=#{v}"
+      end
+
+      host = "eutils.ncbi.nlm.nih.gov"
+      path = "/entrez/eutils/esearch.fcgi?tool=bioruby&db=pubmed&#{opts.join('&')}&term="
+
+      http = Net::HTTP.new(host)
+      response, = http.get(path + CGI.escape(str))
+      result = response.body
+      result = result.scan(/<Id>(.*?)<\/Id>/m).flatten
+      return result
+    end
+
+    def self.efetch(*ids)
+      return [] if ids.empty?
+
+      host = "eutils.ncbi.nlm.nih.gov"
+      path = "/entrez/eutils/efetch.fcgi?tool=bioruby&db=pubmed&retmode=text&rettype=medline&id="
+
+      ids = ids.join(",")
+
+      http = Net::HTTP.new(host)
+      response, = http.get(path + ids)
+      result = response.body
+      result = result.split(/\n\n+/)
+      return result
+    end
+
+  end
+
+end
+
+
+if __FILE__ == $0
+
+  puts Bio::PubMed.query("10592173")
+  puts "--- ---"
+  puts Bio::PubMed.pmfetch("10592173")
+  puts "--- ---"
+  Bio::PubMed.search("(genome AND analysis) OR bioinformatics)").each do |x|
+    p x
+  end
+  puts "--- ---"
+  Bio::PubMed.esearch("(genome AND analysis) OR bioinformatics)").each do |x|
+    p x
+  end
+  puts "--- ---"
+  puts Bio::PubMed.efetch("10592173", "14693808")
+
+end
+
+=begin
+
+= Bio::PubMed
+
+These class methods access NCBI/PubMed database via HTTP.
+
+--- Bio::PubMed.esearch(str, options)
+
+      Search keywords in PubMed by E-Utils and returns an array of PubMed IDs.
+      Options can be a hash containing keys include 'field', 'reldate',
+      'mindate', 'maxdate', 'datetype', 'retstart', 'retmax', 'retmode',
+      and 'rettype' as specified in the following URL:
+
+        ((<URL:http://eutils.ncbi.nlm.nih.gov/entrez/query/static/esearch_help.html#PubMed>))
+
+     Default 'retmax' is 100.
+
+--- Bio::PubMed.efetch(pmids)
+
+      Returns an array of MEDLINE records.  A list of PubMed IDs can be
+      supplied as following:
+
+        Bio::PubMed.efetch(123)
+        Bio::PubMed.efetch(123,456,789)
+        Bio::PubMed.efetch([123,456,789])
+
+--- Bio::PubMed.query(pmid)
+
+      Retrieve PubMed entry by PMID and returns MEDLINE format string (can
+      be parsed by the Bio::MEDLINE and can be converted into Bio::Reference
+      object).
+
+--- Bio::PubMed.pmfetch(pmid)
+
+      Just another query method (by pmfetch).
+
+--- Bio::PubMed.search(str)
+
+      Search the PubMed database by given keywords and returns the list of
+      matched records in MEDLINE format.
+
+
+= For more informations
+
+* Overview
+  * ((<URL:http://www.ncbi.nlm.nih.gov/entrez/query/static/overview.html>))
+* How to link
+  * ((<URL:http://www.ncbi.nlm.nih.gov/entrez/query/static/linking.html>))
+* MEDLINE format
+  * ((<URL:http://www.ncbi.nlm.nih.gov/entrez/query/static/help/pmhelp.html#MEDLINEDisplayFormat>))
+* Search field descriptions and tags
+  * ((<URL:http://www.ncbi.nlm.nih.gov/entrez/query/static/help/pmhelp.html#SearchFieldDescriptionsandTags>))
+* Entrez utilities index
+  * ((<URL:http://www.ncbi.nlm.nih.gov/entrez/utils/utils_index.html>))
+* PmFetch CGI help
+  * ((<URL:http://www.ncbi.nlm.nih.gov/entrez/utils/pmfetch_help.html>))
+* E-Utilities CGI help
+  * ((<URL:http://eutils.ncbi.nlm.nih.gov/entrez/query/static/eutils_help.html>))
+
+=end
+
+

data/lib/bio/io/registry.rb

@@ -0,0 +1,308 @@
+#
+# = bio/io/registry.rb - OBDA BioRegistry module
+#
+# Copyright::   Copyright (C) 2002, 2003, 2004, 2005
+#               Toshiaki Katayama <k@bioruby.org>
+# License::     LGPL
+#
+# $Id: registry.rb,v 1.16 2005/12/18 15:58:42 k Exp $
+#
+# == Description
+#
+# BioRegistry read the OBDA (Open Bio Database Access) configuration file
+# (seqdatabase.ini) and create a registry object.  OBDA is created during
+# the BioHackathon held in Tucson and South Africa in 2002 as a project
+# independent set of protocols to access biological databases.  The spec
+# is refined in the BioHackathon 2003 held in Singapore.
+#
+# By using the OBDA, user can access to the database by get_database method
+# without knowing where and how the database is stored, and each database
+# has the get_by_id method to obtain a sequence entry.
+#
+# Sample configuration file is distributed with BioRuby package which
+# consists of stanza format entries as following:
+#
+#   VERSION=1.00
+#
+#   [myembl]
+#   protocol=biofetch
+#   location=http://www.ebi.ac.uk/cgi-bin/dbfetch
+#   dbname=embl
+#
+#   [mysp]
+#   protocol=biosql
+#   location=db.bioruby.org
+#   dbname=biosql
+#   driver=mysql
+#   user=root
+#   pass= 
+#   biodbname=swissprot
+#
+# The first line means that this configration file is version 1.00.
+#
+# The [myembl] line defines a user defined database name 'myembl' and
+# following block indicates how the database can be accessed.
+# In this example, the 'myembl' database is accecced via the OBDA's
+# BioFetch protocol to the dbfetch server at EBI, where the EMBL
+# database is accessed by the name 'embl' on the server side.
+#
+# The [mysp] line defines another database 'mysp' which accesses the
+# RDB (Relational Database) at the db.bioruby.org via the OBDA's
+# BioSQL protocol.  This BioSQL server is running MySQL database as
+# its backend and stores the SwissProt database by the name 'swissprot'
+# and which can be accessed by 'root' user without password.
+# Note that the db.bioruby.org server is a dummy for the explanation.
+#
+# The configuration file is searched by the following order.
+#
+# 1. Local file name given to the Bio::Registry.new(filename).
+#
+# 2. Remote or local file list given by the environmenetal variable
+#    'OBDA_SEARCH_PATH', which is a '+' separated string of the
+#    remote (HTTP) and/or local files.
+#
+#      e.g. OBDA_SEARCH_PATH="http://example.org/obda.ini+$HOME/lib/myobda.ini"
+#
+# 3. Local file "$HOME/.bioinformatics/seqdatabase.ini" in the user's
+#    home directory.
+#
+# 4. Local file "/etc/bioinformatics/seqdatabase.ini" in the system
+#    configuration directry.
+#
+# All these configuration files are loaded.  If there are database
+# definitions having the same name, the first one is used.
+#
+# If none of these files can be found, Bio::Registry.new will try
+# to use http://www.open-bio.org/registry/seqdatabase.ini file.
+#
+# == References
+# 
+# * http://obda.open-bio.org/
+# * http://cvs.open-bio.org/cgi-bin/viewcvs/viewcvs.cgi/obda-specs/?cvsroot=obf-common
+# * http://www.open-bio.org/registry/seqdatabase.ini
+#
+#--
+#  This library is free software; you can redistribute it and/or
+#  modify it under the terms of the GNU Lesser General Public
+#  License as published by the Free Software Foundation; either
+#  version 2 of the License, or (at your option) any later version.
+#
+#  This library is distributed in the hope that it will be useful,
+#  but WITHOUT ANY WARRANTY; without even the implied warranty of
+#  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+#  Lesser General Public License for more details.
+#
+#  You should have received a copy of the GNU Lesser General Public
+#  License along with this library; if not, write to the Free Software
+#  Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307  USA
+#
+#++
+#
+
+require 'uri'
+require 'net/http'
+
+
+module Bio
+
+autoload :Fetch,          'bio/io/fetch'
+autoload :SQL,            'bio/io/sql'
+autoload :FlatFile,       'bio/io/flatfile'
+autoload :FlatFileIndex,  'bio/io/flatfile/index'
+
+class Registry
+
+  def initialize(file = nil)
+    @spec_version = nil
+    @databases = Array.new
+    read_local(file) if file
+    env_path = ENV['OBDA_SEARCH_PATH']
+    if env_path and env_path.size > 0
+      read_env(env_path)
+    else
+      read_local("#{ENV['HOME']}/.bioinformatics/seqdatabase.ini")
+      read_local("/etc/bioinformatics/seqdatabase.ini")
+      if @databases.empty?
+        read_remote("http://www.open-bio.org/registry/seqdatabase.ini")
+      end
+    end
+  end
+
+  # Version string of the first configulation file
+  attr_reader :spec_version
+
+  # List of databases (Array of Bio::Registry::DB)
+  attr_reader :databases
+
+  # Returns a dababase handle (Bio::SQL, Bio::Fetch etc.) or nil
+  # if not found (case insensitive).
+  # The handles should have get_by_id method.
+  def get_database(dbname)
+    @databases.each do |db|
+      if db.database == dbname.downcase
+        case db.protocol
+        when 'biofetch'
+          return serv_biofetch(db)
+        when 'biosql'
+          return serv_biosql(db)
+        when 'flat', 'index-flat', 'index-berkeleydb'
+          return serv_flat(db)
+        when 'bsane-corba', 'biocorba'
+          raise NotImplementedError
+        when 'xembl'
+          raise NotImplementedError
+        end
+      end
+    end
+    return nil
+  end
+  alias db get_database
+
+  # Returns a Registry::DB object corresponding to the first dbname
+  # entry in the registry records (case insensitive).
+  def query(dbname)
+    @databases.each do |db|
+      return db if db.database == dbname.downcase
+    end
+  end
+
+  private
+
+  def read_env(path)
+    path.split('+').each do |elem|
+      if /:/.match(elem)
+        read_remote(elem)
+      else
+        read_local(elem)
+      end
+    end
+  end
+
+  def read_local(file)
+    if File.readable?(file)
+      stanza = File.read(file)
+      parse_stanza(stanza)
+    end
+  end
+
+  def read_remote(url)
+    schema, user, host, port, reg, path, = URI.split(url)
+    Net::HTTP.start(host, port) do |http|
+      response, = http.get(path)
+      parse_stanza(response.body)
+    end
+  end
+
+  def parse_stanza(stanza)
+    return unless stanza
+    if stanza[/.*/] =~ /VERSION\s*=\s*(\S+)/
+      @spec_version ||= $1	# for internal use (may differ on each file)
+      stanza[/.*/] = ''	        # remove VERSION line
+    end
+    stanza.each_line do |line|
+      case line
+      when /^\[(.*)\]/
+        dbname = $1.downcase
+        db = Bio::Registry::DB.new($1)
+        @databases.push(db)
+      when /=/
+        tag, value = line.chomp.split(/\s*=\s*/)
+        @databases.last[tag] = value
+      end
+    end
+  end
+
+  def serv_biofetch(db)
+    serv = Bio::Fetch.new(db.location)
+    serv.database = db.dbname
+    return serv
+  end
+
+  def serv_biosql(db)
+    location, port = db.location.split(':')
+    port = db.port unless port
+
+    case db.driver
+    when /mysql/i
+      driver = 'Mysql'
+    when /pg|postgres/i
+      driver = 'Pg'
+    when /oracle/
+    when /sybase/
+    when /sqlserver/
+    when /access/
+    when /csv/
+    when /informix/
+    when /odbc/
+    when /rdb/
+    end
+
+    dbi = [ "dbi", driver, db.dbname, location ].compact.join(':')
+    dbi += ';port=' + port if port
+    serv = Bio::SQL.new(dbi, db.user, db.pass)
+
+    # We can not manage biodbname (for name space) in BioSQL yet.
+    # use db.biodbname here!!
+
+    return serv
+  end
+
+  def serv_flat(db)
+    path = db.location
+    path = File.join(path, db.dbname) if db.dbname
+    serv = Bio::FlatFileIndex.open(path)
+    return serv
+  end
+
+
+  class DB
+
+    def initialize(dbname)
+      @database = dbname
+      @property = Hash.new
+    end
+    attr_reader :database
+
+    def method_missing(meth_id)
+      @property[meth_id.id2name]
+    end
+
+    def []=(tag, value)
+      @property[tag] = value
+    end
+
+  end
+
+end # class Registry
+
+end # module Bio
+
+
+
+if __FILE__ == $0
+  begin
+    require 'pp'
+    alias p pp
+  rescue
+  end
+
+  # Usually, you don't need to pass ARGV.
+  reg = Bio::Registry.new(ARGV[0])
+
+  p reg
+  p reg.query('genbank_biosql')
+
+  serv = reg.get_database('genbank_biofetch')
+  puts serv.get_by_id('AA2CG')
+
+  serv = reg.get_database('genbank_biosql')
+  puts serv.get_by_id('AA2CG')
+
+  serv = reg.get_database('swissprot_biofetch')
+  puts serv.get_by_id('CYC_BOVIN')
+
+  serv = reg.get_database('swissprot_biosql')
+  puts serv.get_by_id('CYC_BOVIN')
+end
+
+