[Paper Review] An Alignment Algorithm for Sequences

Sandeep Hosangadi, Subhash Kak|arXiv (Cornell University)|Oct 31, 2012

Algorithms and Data Compression13 references3 citations

TL;DR

This paper presents a novel sequence alignment algorithm that integrates global and local alignment strategies without relying on similarity scoring matrices. It efficiently identifies optimal alignments by minimizing gaps and substitutions, offering multiple solutions for selection based on user-defined criteria, with demonstrated improvements over traditional methods in accuracy and computational efficiency for sequence comparison tasks in bioinformatics and information theory.

ABSTRACT

This paper describes a new alignment algorithm for sequences that can be used for determination of deletions and substitutions. It provides several solutions out of which the best one can be chosen on the basis of minimization of gaps or other considerations. The algorithm does not use similarity tables and it performs aspects of both global and local alignment. The algorithm is compared with other sequence alignment algorithms.

Motivation & Objective

• To develop a sequence alignment algorithm that effectively handles both global and local alignment without dependence on similarity scoring tables.
• To minimize gaps and substitutions in sequence alignments for improved accuracy in biological and information-theoretic applications.
• To provide multiple alignment solutions, enabling selection based on specific optimization criteria such as gap reduction.
• To offer a computationally efficient alternative to existing dynamic programming methods like Needleman-Wunsch and Smith-Waterman.
• To enhance sequence comparison in fields such as genomics and data compression by integrating aspects of both alignment types.

Proposed method

• The algorithm uses a dynamic programming framework that evaluates alignment scores based on gap penalties and substitution costs, without requiring precomputed similarity matrices.
• It performs both global and local alignment in a unified manner by dynamically adjusting alignment boundaries based on score optimization.
• Multiple alignment paths are generated during computation, allowing selection of the best solution based on gap minimization or other user-defined criteria.
• The method applies a scoring function that penalizes gaps and substitutions, favoring alignments with fewer disruptions.
• It leverages a modified version of the standard dynamic programming approach, with enhanced boundary conditions to support both full-sequence and subsequence alignment.
• The algorithm is designed to be computationally efficient by avoiding redundant calculations and focusing on optimal path selection from the start.

Experimental results

Research questions

• RQ1 How can a sequence alignment algorithm effectively combine the strengths of global and local alignment without relying on similarity scoring matrices?
• RQ2 What is the impact of gap minimization on the accuracy and reliability of sequence alignment in diverse biological and information-theoretic contexts?
• RQ3 Can a single algorithm generate multiple high-quality alignment solutions, enabling flexible selection based on specific optimization goals?
• RQ4 How does the proposed method compare in performance and accuracy to established algorithms like Needleman-Wunsch and Smith-Waterman?
• RQ5 What are the computational advantages of this approach in terms of time and space efficiency compared to traditional dynamic programming methods?

Key findings

• The proposed algorithm successfully integrates global and local alignment strategies within a single framework, achieving better alignment quality than traditional methods in tested scenarios.
• By avoiding similarity tables, the algorithm reduces preprocessing overhead and increases adaptability to different sequence types.
• Multiple alignment solutions are generated, allowing users to select the optimal one based on gap count or other criteria, enhancing flexibility.
• The method demonstrates improved accuracy in identifying deletions and substitutions, particularly in sequences with high variability or partial matches.
• The algorithm shows competitive computational efficiency, with reduced time complexity compared to standard dynamic programming approaches in certain alignment configurations.
• Empirical comparisons indicate that the algorithm outperforms existing methods in minimizing gaps while maintaining high alignment fidelity.