ParallelSubSearch

ParallelSubSearch is a Python project designed to illustrate and explore the algorithmic design and parallelization of functions to find the longest contiguous sub-sequence within a sequence. This project includes a performance comparison between serial and parallel execution approaches.

Description

The project consists of two main parts:

1. Algorithmic Design: Implementing functions longest_contiguous and all_longest to find specific types of sub-sequences within provided data.
2. Parallel Execution: Enhancing performance by implementing a parallel approach to process large datasets using Python's concurrent.futures module.

Getting Started

Dependencies

• Python 3.7 or higher
• pytest for running unit tests

Installing

• Clone the repository to your local machine:

  git clone https://github.com/sapirmardan/ParallelSubSearch.git
  

• Install the required packages:

  pip install numpy biopython pytest
  

Executing Program

• To run the tests and validate functionality:

  python3 -m pytest --doctest-modules -v

• To execute the program and compare the serial vs parallel performance:

  python3 longest_sub.py
  

Implementation Details

Functions:

• longest_contiguous_version1(seq, stop_codon="", gap="", case_sensitive=True): Calculates the longest contiguous sub-sequence for each unique letter in the sequence. It is case-sensitive by default.
• longest_contiguous_version2(seq, stop_codon="", gap="", case_sensitive=True): Similar to version 1 but optimized for performance using NumPy for large sequences.
• all_longest: This function finds all sub-sequences which are longest by length and not necessarily contiguous.
• run_large_test_serial: Executes the data processing in a serial manner.
• run_test_parallel: Executes the data processing in parallel using concurrent.futures to improve performance.

Files

• longest_sub.py: Contains the main logic for sub-sequence identification.
• test_longest_sub.py: Contains unit tests for the functions implemented in longest_sub.py.

Example

Here is a quick example of how to use longest_contiguous:

from longest_sub import longest_contiguous_version1
sequence = Seq("AACCGGTTAACCGGTT")
result = longest_contiguous_version1(sequence)
print(result)