/bio-unidip-examples

Suplementary materials to "Unimodality Based Clustering Applied to Biological Motif Discovery"

Primary LanguageJupyter Notebook

UniDip Testing Examples

This repository archives the suplementary materials for the paper "Unimodality Based Clustering Applied to Biological Motif Discovery." The paper may be viewed at the following link here.

Cloning Repo

git clone https://github.com/BenjaminDoran/bio-unidip-examples.git
git submodule init
git submodule update

Setup

Python

We recommend using the Anaconda 3.6 version of Python with these packages:

BioPython Numpy Pandas Matplotlib

MUSCLE

Go to https://www.drive5.com/muscle/ to download the MUSCLE alignment tool for your platform. unzip the folder and place the executable in the base folder of this repo. All notebooks that use MUSCLE have a variable that holds the path to the executable.

MEME

runmeme.sh starts and runs a docker conatiner with meme installed. The file also contains an example command on how to run MEME on the raw sequences. Adjust the input file and output folder to run on trimmed sequences.

Examples

  • numeric.ipynb provides examples of generated random numeric samples.
  • letters.ipynb provides examples looking at generated symbolic data.
  • aligned.ipynb provides examples of isolating motifs after global alignments.
  • real_data.ipynb retrives sequences from JASPER and NCBI
  • real_data_align.ipynb provides examples of UniDip applied to real genomic sequences.
  • util.py provides helper functions for data transforms and string generation.