Skip to content
/ BioDSL Public

BioDSL (pronounced Biodiesel) is a Domain Specific Language for creating bioinformatic analysis workflows.

License

GPL-2.0 license
16 stars 3 forks Branches Tags Activity
Star
Notifications You must be signed in to change notification settings

maasha/BioDSL

BranchesTags

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

1,114 Commits
examples
examples
 
 
lib
lib
 
 
test
test
 
 
www
www
 
 
.gitignore
.gitignore
 
 
BioDSL.gemspec
BioDSL.gemspec
 
 
Gemfile
Gemfile
 
 
LICENSE
LICENSE
 
 
README.md
README.md
 
 
Rakefile
Rakefile
 
 

Repository files navigation

BioDSL (pronounced Biodiesel) is a Domain Specific Language for creating bioinformatic analysis workflows. A workflow may consist of several pipelines and each pipeline consists of a series of steps such as reading in data from a file, processing the data in some way, and writing data to a new file.

BioDSL is build on the same principles as Biopieces, where data records are passed through multiple commands each with a specific task. The idea is that a command will process the data record if this contains the relevant attributes that the command can process. E.g. if a data record contains a sequence, then the command reverse_seq will reverse that sequence.

Installation

The recommended way of installing BioDSL is via Ruby’s gem package manager:

$ gem install BioDSL

For those commands which are wrappers around third-party tools, such as Usearch, Mothur and SPAdes, you will have to install these and make the executables available in your $PATH.

Getting started

BioDSL is implemented in Ruby making use of Ruby’s powerful metaprogramming facilities. Thus, a workflow is basically a Ruby script containing one or more pipelines.

Here is a test script with a single pipeline that reads all FASTA entries from the file input.fna, selects all records with a sequence ending in ATC, and writing those records as FASTA entries to the file output.fna:

#!/usr/bin/env ruby

require 'BioDSL'

BD.new.
read_fasta(input: "input.fna").
grab(select: "ATC$", keys: :SEQ).
write_fasta(output: "output.fna").
run

Save the test script to a file test.biodsl and execute on the command line:

$ ruby test.biodsl

Combining multiple pipelines

This script demonstrates how multiple pipelines can be created and combined. In the end two pipelines are run, one consisting of p1 + p2 and one consisting of p1 + p3. The first pipeline run will produce a histogram plot of sequence length from sequences containing the pattern ATCG, and the other pipeline run will produce a plot with sequences length distribution of sequences not matching ATCG.

#!/usr/bin/env ruby

require 'BioDSL'

p1 = BD.new.read_fasta(input: "test.fna")
p2 = BD.new.grab(keys: :SEQ, select: "ATCG").
     plot_histogram(key: :SEQ_LEN, terminal: :png, output: "select.png")
p3 = BD.new.grab(keys: :SEQ, reject: "ATCG").
     plot_histogram(key: :SEQ_LEN, terminal: :png, output: "reject.png")
p4 = p1 + p3

(p1 + p2).write_fasta(output: "select.fna").run
p4.write_fasta(output: "reject.fna").run

Running pipelines in parallel

This script demonstrates how to run multiple pipelines in parallel using 20 CPU cores. Here we filter pair-end FASTQ entries from a list of samples described in the file samples.txt which contains three tab separated columns: sample name, a forward read file path, and a reverse read file path.

#!/usr/bin/env ruby

require 'BioDSL'
require 'csv'

samples = CSV.read("samples.txt")

Parallel.each(samples, in_processes: 20) do |sample|
  BD.new.
  read_fastq(input: sample[1], input2: sample[2], encoding: :base_33).
  grab(keys: :SEQ, select: "ATCG").
  write_fastq(output: "#{sample[0]}_filted.fastq.bz2", bzip2: true).
  run
end

Ruby one-liners

It is possible to execute BioDSL pipelines on the command line:

ruby -r BioDSL -e 'BD.new.read_fasta(input: "test.fna").plot_histogram(key: :SEQ_LEN).run'

And to save typing we may use the alias bd which is set like this on the command line:

$ alias bd='ruby -r BioDSL'

It may be a good idea to save that alias in your .bashrc file.

Now it is possible to run a BioDSL pipeline on the command line like this:

$ bd -e 'BD.new.read_fasta(input: "test.fna").plot_histogram(key: :SEQ_LEN).run'

Using the Interactive Ruby interpreter

Here we demonstrate the use of Ruby's irb shell:

$ irb -r BioDSL --noinspect
irb(main):001:0> p = BD.new
=> BD.new
irb(main):002:0> p.read_fasta(input: "input.fna")
=> BD.new.read_fasta(input: "input.fna")
irb(main):003:0> p.grab(select: "ATC$", keys: :SEQ)
=> BD.new.read_fasta(input: "input.fna").grab(select: "ATC$", keys: :SEQ)
irb(main):004:0> p.write_fasta(output: "output.fna")
=> BD.new.read_fasta(input: "input.fna").grab(select: "ATC$", keys: :SEQ).write_fasta(output: "output.fna")
irb(main):005:0> p.run
=> BD.new.read_fasta(input: "input.fna").grab(select: "ATC$", keys: :SEQ).write_fasta(output: "output.fna").run
irb(main):006:0>

Again, it may be a good idea to save an alias alias biodsl="irb -r BioDSL --noinspect" to your .bashrc file. Thus, we can use the new biodsl alias to chain commands directly:

$ biodsl
irb(main):001:0> BD.new.read_fasta(input: "input.fna").grab(select: "ATC$", keys: :SEQ).write_fasta(output: "output.fna").run(progress: true)
=> BD.new.read_fasta(input: "input.fna").grab(select: "ATC$", keys: :SEQ).write_fasta(output: "output.fna").run(progress: true)
irb(main):002:0>

History file

A history file is kept in $USER/.BioDSL_history and each time run is called a history entry is added to this file:

BD.new.read_fasta(input: "test_big.fna", first: 100).plot_histogram(key: :SEQ_LEN).run
BD.new.read_fasta(input: "test_big.fna", first: 100).plot_histogram(key: :SEQ_LEN).run
BD.new.read_fasta(input: "test_big.fna", first: 10).plot_histogram(key: :SEQ_LEN).run
BD.new.read_fasta(input: "test_big.fna").plot_histogram(key: :SEQ_LEN).run
BD.new.read_fasta(input: "test_big.fna", first: 1000).plot_histogram(key: :SEQ_LEN).run

Thus it is possible to redo the last pipeline by pasting the line in irb or a Ruby one-liner.

Log and History

All BioDSL events are logged to ~/.BioDSL_log.

BioDSL history is saved to ~/.BioDSL_history.

Features

Progress

Show nifty progress table with commands, records read and emittet and time.

BD.new.read_fasta(input: "input.fna").dump.run(progress: true)

Verbose

Output verbose messages from commands and the run status.

BD.new.read_fasta(input: "input.fna").dump.run(verbose: true)

Debug

Output debug messages from commands using these.

BD.new.read_fasta(input: "input.fna").dump.run(debug: true)

E-mail notification

Send an email when run is complete.

BD.new.read_fasta(input: "input.fna").dump.run(email: [email protected], subject: "Script done!")

Reports

Create an HTML report of the run stats for a pipeline:

BD.new.read_fasta(input: "input.fna").dump.run(report: "status.html")

Output directory

All output files from commands are put in a specified directory:

BD.new.read_fasta(input: "input.fna").dump.run(output_dir: "Results")

Configuration File

It is possible to pre-set options in a configuration file located in your $HOME directory called .BioDSLrc. Thus if an option is not already set, its value will fall back to the one set in the configuration file. The configuration file contains three whitespace separated columns:

  • Command name
  • Option
  • Option value

Lines starting with # are considered comments and are ignored.

An example:

maasha@mel:~$ cat ~/.BioDSLrc
uchime_ref   database   /home/maasha/Install/QIIME1.8/data/rdp_gold.fa
uchime_ref   cpus       20

On compute clusters it is necessary to specify the max processor count, which is otherwise determined as the number of cores on the current node. To override this add the following line:

pipeline   processor_count   1000

It is also possible to change the temporary directory from the systems default by adding the following line:

pipeline   tmp_dir   /home/projects/ku_microbio/scratch/tmp

Available BioDSL commands

Running the test suite

BioDSL have an extended set of unit tests that can be run after installing development dependencies. First you need to install the bundler gem:

$ gem install bundler

Next you need to change to the source directory of BioDSL and run bundler to download depending gems:

$ bundle install

And then you run the test suite by running rake:

$ rake

And the unit tests should all run, except those omitted because a third-party executable was missing.

Contributing

  1. Fork it
  2. Create your feature branch (git checkout -b my-new-feature)
  3. Commit your changes (git commit -am 'Add some feature')
  4. Push to the branch (git push origin my-new-feature)
  5. Create new Pull Request

About

BioDSL (pronounced Biodiesel) is a Domain Specific Language for creating bioinformatic analysis workflows.

Resources

Readme

License

GPL-2.0 license
Activity

Stars

16 stars

Watchers

4 watching

Forks

3 forks
Report repository

Releases

34 tags

Packages

No packages published

Languages