A brief and unscientific survey of social bookmarking buttons provided by journal websites. Take with pinch of salt and/or tongue in cheek.

PLoS get it. As you'd expect.	BMC get it too.	Even ACS kind of get it.
OUP get it to a degree.	Someone doesn't get it.

I know, all bookmarking services provide their own tools, such as bookmarklets. However, I suspect that a button is more convenient for many people: it’s a courtesy and an advertisement for social bookmarking and online reference management tools.

So NPG: is it “we promote sharing” or “we promote sharing so long as you use our service?”

Another year, another ISMB/ECCB meeting and – another great blogging effort.

It’s all at the FriendFeed group: ISMB/ECCB Stockholm 2009, with outgoing links to individual blogs too.

Thanks and congratulations to all involved for a great effort. Looking forward to the official write-up.

I’m in the process of adding search to a Rails application, using acts_as_ferret and ran into this issue. How to search a table using a field from another table, supplied by a foreign key?

Let’s say you created a table for a model named Feature:

  create_table "features", :force => true do |t|
    t.integer  "feature_name_id"
    t.integer  "chromosome_id"
    t.integer  "start"
    t.integer  "end"
    t.string   "strand"
    t.datetime "created_at"
    t.datetime "updated_at"
  end

This model, Feature, belongs_to two models, FeatureName and Chromosome, each with a field “name”:

  create_table "feature_names", :force => true do |t|
    t.string   "name"
    t.datetime "created_at"
    t.datetime "updated_at"
  end

  create_table "chromosomes", :force => true do |t|
    t.string   "name"
    t.integer  "length"
    t.datetime "created_at"
    t.datetime "updated_at"
  end

You want to search for features using the associated feature_name or chromosome_name.

Here’s the trick. Edit app/models/feature.rb to look something like this (just showing the relevant parts here):

class Feature < ActiveRecord::Base
  acts_as_ferret :fields => [:fname,:cname]
  belongs_to :feature_name
  belongs_to :chromosome

  # ferret indexing on foreign keys
  def fname
    "#{self.feature_name.name}"
  end

  def cname
    "#{self.chromosome.name}"
  end

end

That tells acts_as_ferret to index whatever is returned by the methods fname and cname. Which is: the name field from the FeatureName and Chromosome models, respectively.

For testing, I use a small rake task to pre-build indices. Run it with “rake ferret:build”.

namespace :ferret do
  # Rebuild index task.
    desc "Rebuild a Ferret index."
      task :build => :environment do
        [Feature,MyOtherModel1,MyOtherModel2].each do |model|
            model.rebuild_index
        end
  end
end

If search isn’t returning results following your changes, don’t forget to restart the server.

I’ve been banging my head against the wall for almost a week with a Rails application. This post is not a plea for help – I’d use a forum for that – just a record of the problem. That said, feel free to comment, especially if you have a similar problem.

This is all using Rails 2.3.2, Mongrel 1.1.5, installed as gems on Ubuntu 9.04.

The basic issue: 2 models, 2 controllers, 2 sets of views. Identical in almost every respect, little more than basic CRUD (index, create, update, destroy). (1) works, (2) does not.

Update – thanks for your comments, here and elsewhere. In the end I rebuilt from scratch with scaffolding and it’s all good so far. Guess there was something rogue in my hand-crafted code.

The details:
1. The models
Model Organism:

class Organism < ActiveRecord::Base
  has_many :chromosomes
  has_many :platforms

  validates_presence_of     :name,:binomial,:taxid
  validates_numericality_of :taxid
  validates_uniqueness_of   :name,:binomial,:taxid
end

Model Chromosome:

class Chromosome < ActiveRecord::Base
  belongs_to :organism
  has_many   :features

  validates_presence_of     :name, :length
  validates_numericality_of :length
  validates_uniqueness_of   :name, :scope => :organism_id
end

2. The routes
File routes.rb contains (just the relevant lines):

ActionController::Routing::Routes.draw do |map|
  map.resources :chromosomes
  map.resources :organisms
end

3. Controller for Chromosome (this works)
Here’s the controller for Chromosome. It’s just your basic CRUD.
chromosomes_controller.rb

class ChromosomesController < ApplicationController

  def index
    @chromosomes = Chromosome.all
  end

  def edit
    @chromosome = Chromosome.find(params[:id])
  end

  def update
    @chromosome = Chromosome.find(params[:id])
    if @chromosome.update_attributes(params[:chromosome])
      flash[:notice] = "Chromosome updated"
      redirect_to :action => :index
    else
      render :action => :new
    end
  end

  def new
    @chromosome = Chromosome.new
  end

  def create
    @chromosome = Chromosome.new(params[:chromosome])
    if @chromosome.save
      flash[:notice] = "Chromosome added"
      redirect_to :action => "index"
    else
      render :action => "new"
    end
  end

  def destroy
    @chromosome = Chromosome.find(params[:id])
    # don't delete chromosome if linked to features
    if Feature.all(:conditions => {:chromosome_id => @chromosome.id}).length > 0
      flash[:notice] = "Cannot delete chromosome:  it has features"
      redirect_to :action => "index"
    else
      @chromosome.destroy
      flash[:notice] = "Chromosome deleted"
      redirect_to :action => "index"
    end
  end

end

3. Views for Chromosome (these work too)
Now the views: index, edit, new and a partial named “_form.html.erb”.
index.html.erb:

<h1>Chromosomes</h1>
<div class="info">Add a <%= link_to 'new chromosome', new_chromosome_path %> here.</div>
<div class="info">Or click a 'last update' table entry to view/edit/delete.</div>

<table>
 <tr>
   <th>Name</th>
   <th>Length</th>
   <th>Organism</th>
   <th>Last Updated</th>
 </tr>

 <% @chromosomes.each do |@chrom| %>
   <tr class="data">
     <td><%=h @chrom.name %></td>
     <td><%=h @chrom.length %></td>
     <td><%=h @chrom.organism.name %></td>
     <td><%= link_to @chrom.updated_at, edit_chromosome_path(@chrom) %></td>
   </tr>
 <% end %>
</table>

new.html.erb and edit.html.erb – are essentially identical except for wording:
new.html.erb

<h1>New Sequence</h1>
<div class="info">Enter sequence details and click 'Save'.</div>

<%= render :partial => "form" %>

Here’s the partial ERB file, _form.html.erb:

<% form_for(@chromosome) do |f| %>
  <%= f.error_messages %>
  <table>
    <tr>
      <td><%= f.label      :name %></td>
      <td><%= f.text_field :name %></td>
    </tr>
    <tr>
      <td><%= f.label      :length %></td>
      <td><%= f.text_field :length %></td>
    </tr>
    <tr>
      <td><%= f.label      :organism %></td>
      <td>
        <%= render :partial => "organisms/select", :locals => {:organisms => @organisms} %>
      </td>
    </tr>
    <tr>
      <td colspan="2" class="right"><%= f.submit 'Save' %></td>
    </tr>
  </table>
<% end %>

<% if @chromosome.id %>
<div class="info">Or <%= link_to 'delete', @chromosome,
                         :confirm => "Are you sure?",
                         :method => :delete %> this chromosome permanently.</div>
<% end %>

The partial app/views/organisms/_select.html.erb is simply:

<%= collection_select(:chromosome, :organism_id, Organism.all, :id, :name) %>

4. Controller and views for Organism (these don’t work)
The controller organisms_controller.rb is identical to that for chromosomes, EXCEPT that @chromosome(s) is replaced by @organism(s) throughout and the destroy method is:

  def destroy
    @organism = Organism.find(params[:id])
    # don't delete organism if linked to chromosomes
    if Chromosome.all(:conditions => {:organism_id => @organism.id}).length > 0
      flash[:notice] = "Cannot delete organism:  it has chromosomes"
      redirect_to :action => "index"
    else
      @organism.destroy
      flash[:notice] = "Organism deleted"
      redirect_to :action => "index"
    end
  end

And index.html.erb, new.html.erb, edit.html.erb are also identical, except that @chrom(osome) is replaced by @org(anism) throughout and the fields are those for Organism. For example in _form.html.erb:

<% form_for(@organism) do |f| %>
  <%= f.error_messages %>
  <table>
    <tr>
      <td><%= f.label      :name %></td>
      <td><%= f.text_field :name %></td>
    </tr>
    <tr>
      <td><%= f.label      :binomial %></td>
      <td><%= f.text_field :binomial %></td>
    </tr>
    <tr>
      <td><%= f.label      :taxid %></td>
      <td><%= f.text_field :taxid %></td>
    </tr>
    <tr>
      <td colspan="2" class="right"><%= f.submit 'Save' %></td>
    </tr>
  </table>
<% end %>

<% if @organism.id %>
<div class="info">Or <%= link_to 'delete', @organism,
                         :confirm => "Are you sure?",
                         :method => :delete %> this organism permanently.</div>
<% end %>

5. The problem
Chromosome works just fine. Index shows a list of chromosomes. New adds a new chromosome. Edit fills in the form with chromosome parameters and allows update or destroy.

Organism does not work just fine. The new method works and adds a new organism. But edit gives this error (I removed the ERB/HTML tags as they mess up blog formatting):

NoMethodError in Organisms#edit
Showing app/views/organisms/_form.html.erb where line #1 raised:
undefined method `organism__path' for #ActionView::Base:0xb6a81fe0
Extracted source (around line #1):

/usr/lib/ruby/gems/1.8/gems/actionpack-2.3.2/lib/action_controller/polymorphic_routes.rb:109:in `__send__'
/usr/lib/ruby/gems/1.8/gems/actionpack-2.3.2/lib/action_controller/polymorphic_routes.rb:109:in `polymorphic_url'
/usr/lib/ruby/gems/1.8/gems/actionpack-2.3.2/lib/action_controller/polymorphic_routes.rb:116:in `polymorphic_path'
/usr/lib/ruby/gems/1.8/gems/actionpack-2.3.2/lib/action_view/helpers/form_helper.rb:298:in `apply_form_for_options!'
/usr/lib/ruby/gems/1.8/gems/actionpack-2.3.2/lib/action_view/helpers/form_helper.rb:277:in `form_for'
code/Rails/ArrayMiner/app/views/organisms/_form.html.erb:1:in `_run_erb_app47views47organisms47_form46html46erb_locals_form_object'
code/Rails/ArrayMiner/app/views/organisms/edit.html.erb:4:in `_run_erb_app47views47organisms47edit46html46erb'


The problem?
There seems to be some issue with routes. For some reason, form_for(@organism) works fine with the new method, but not with the edit method. The error “undefined method ‘organism__path’” (note the 2 underscores) is perplexing.
Other observations – (1) moving lines around in routes.rb then restarting Mongrel has sporadically fixed the problem, but then errors return. (2) I have similar code for a model named Technology which misbehaves in the same way, but code for 3 other models in addition to Chromosome works fine. As before, all the code is more or less identical, just with different variable names.

Here’s a problem. You’d like to construct a complex query at NCBI Entrez using various fields. Example:

“9606″[Taxonomy ID]

to limit your search to Homo sapiens. Except – you don’t know which fields are available for the database that you want to query.

EInfo can return an XML file with this information. Ruby + Hpricot eats XML for breakfast. Here’s an example using the GEO Datasets (gds) database.

#!/usr/bin/ruby

require 'rubygems'
require 'hpricot'
require 'open-uri'

doc = Hpricot(open("http://eutils.ncbi.nlm.nih.gov/entrez/eutils/einfo.fcgi?db=gds"))

(doc/'//fieldlist/field').each do |f|
  puts "#{(f/'/name').inner_html},#{(f/'/fullname').inner_html},#{(f/'description').inner_html}"
end

And the first few lines of output:

ALL,All Fields,All terms from all searchable fields
UID,UID,Unique number assigned to publication
FILT,Filter,Limits the records
ORGN,Organism,exploded organism names
....24 more lines....


Posted in bioinformatics, computing, ruby Tagged: entrez, eutils, hpricot, ncbi ]]>