Pairwise sequence alignment using Hidden Markov Models

Project repository for Bioinformatics course at Faculty of Electrical Engineering and Computing, University of Zagreb, 2018./2019.

Description

Main goal of this project is implementation of an algorithm for pairwise sequence alignment using HMM.

• Project documentation
• Presentation

Authors

• Tomislav Božurić
• Martin Pisačić
• Krešimir Topolovec

Installation

Clone this repo and go to ./build directory

git clone git@github.com:Sokre95/bioinf_project.git
cd bioinf_project/build

Build bioinf executable

cmake ../
make

Usage

Execute ./bioinf --help to print command info

Run either with -v [--viterbi] or -e [--estimate] option. Both options can't be used at the  same time
Usage:
  bioinf [OPTION...]

 OPTIONS options:
  -v, --viterbi arg           # Run sequence alignment algorithm on sequence pair given in fasta file 
                                specified by <arg> path
  -e, --estimate arg          # Run HMM parameters estimator with path to directory holding learning database
                                specified by  <arg> path. Calculated parameters are stored in ./params.txt file
  -o, --out arg               # [Use only with -v option] Write aligned sequences to ./aligned/{pair_file_name}.fasta
                                (default: true)
  -c, --console               # [Use only with -v option] Print aligned sequences to console
  -m, --multiline [=N(=100)]  # [Use only with -v option] Write/Print aligned sequences in multiple lines, each line
                                containg N chars
  -p, --progress              # Show progress while running algorithm
  -t, --mem-optimized         # Run memory optimized version of Viterbi algorithm
  -h, --help                  # Show help

Input

All input files must be in FASTA format and can be placed in arbitrary location. You always have to specify path to the file or folder you are using.

Examples

Run sequence alignment for p2.fasta file writing 200 chars in each line of output file and print output to console also

./bioinf --viterbi ../database/pairs/hepatitis/p2.fasta --console --multiline=200 --progress

Estimate HMM parameters (probabilities) using outputs_mafft/upcase as learning database

./bioinf --estimate ../database/outputs_mafft/upcase

Project structure

All c++ files are located in /src and /include folder. In /helpers folder are small python and ruby scripts used as a help in development of this project.

License

MIT License

2018 Tomislav Božurić, Krešimir Topolovec & Martin Pisačić