Bio-informatic Algorithms

The continuing quest to learn algorithms and data structure used in bioinformatics, and their implementation in Haskell.

Haskell Stack : How to build a project with Stack

Stack is a cross-platform program for developing Haskell projects. It is aimed at Haskellers both new and experienced. Stack can be used to :

• Install GHC automatically, in an isolated location.
• Install packages needed for your project.
• Build your project.
• Test your project.
• Benchmark your project.

Start your new project :

Create a new directory containing all the needed files to start a project,

stack new my-project 
cd my-project

Download the compiler if needed, in an isolated location (defaults to `~/.stack`). The isolated location will not interfere with any system-level installation. Use stack path for information on installation paths.

To build a minimal project :

stack build

To execute :

stack exec my-project-exe

To install an executable use stack,

stack install <package-name>

To launch a REPL :

stack ghci

Project Structure

The project has the following directory structure:

• data to hold the (test) data
• src to hold the source code
• test to hold the test code
• app to hold Main

Under src we have a lib to hold the library :

• Util to hold general utilities, for example Command.hs used for CLI.
• Genome to hold genomic functionality
• Examples to hold examples
• Chapters to hold chapter by chapter code examples

Test code has (almost) identical directory lay-out.

I am stuck with the same ideas, have not been able to break new ground.

Plotting in Haskell

Plotting is essential in any scientific field. However Haskell is not particularly helpful in producing interactive plots. We can take this as an opportunity in scripting our plots --- this requires learning more about Haskell charts. Interactive plotting readily becomes hectic hacking away to get the plot right. We can do the same if we have to script the plot -- however we will have to think more and the end-result of our efforts will be preserved in the script.

TODO <2017-01-06 Fre> center the GC diff plot around its minima.

Faster pattern finding

While we have used dictionaries (Haskell Map) to implement pattern finding algorithms, Compeau and Pevzner suggest using Frequency Arrays. Frequency Arrays work with mapping the four (can be five if 'N' is allowed) nucleotides to numbers, and using a positional factor to convert strings of nucleotides to numbers. For a pattern of length k, there are 4^k possible strings, and hence the numbers will range from 0 to 4^k − 1. This number will be prohibitively large even for moderate values of k. However for the problem of locating the origin of replication, k is typically smaller than 10, and the frequency array will be of the order of a million. We can implement the same pattern search algorithms using the frequency array under their own module.