[Paper Review] Maximum Match Subsequence Alignment Algorithm Finely Grained (MMSAA FG)

Bharath Reddy, Richard Fields|arXiv (Cornell University)|Apr 29, 2023

Algorithms and Data Compression4 citations

TL;DR

This paper introduces MMSAA FG, a novel sequence alignment algorithm that enhances sensitivity in sequence comparison by leveraging adaptive and finely grained perfect match seeds within a suffix tree framework. Building on prior heuristic methods like MASAA, it improves alignment accuracy on long sequences (up to 500k) without sacrificing performance, demonstrating strong results on both synthetic and biological datasets (Rosetta dataset).

ABSTRACT

Sequence alignment is common nowadays as it is used in many fields to determine how closely two sequences are related and at times to see how little they differ. In computational biology / Bioinformatics, there are many algorithms developed over the course of time to not only align two sequences quickly but also get good laboratory results from these alignments. The first algorithms developed were based of a technique called Dynamic Programming, which were very slow but were optimal when it comes to sensitivity. To improve speed, more algorithms today are based of heuristic approach, by sacrificing sensitivity. In this paper, we are going to improve on a heuristic algorithm called MASAA (Multiple Anchor Staged Local Sequence Alignment Algorithm) and MASAA Sensitive which we published previously. This new algorithm appropriately called Maximum Match Subsequence Alignment Algorithm Finely Grained. The algorithm is based on suffix tree data structure like our previous algorithms, but to improve sensitivity, we employ adaptive seeds, and finely grained perfect match seeds in between the already identified anchors. We tested this algorithm on a randomly generated sequences, and Rosetta dataset where the sequence length ranged up to 500 thousand.

Motivation & Objective

• To improve the sensitivity of heuristic sequence alignment algorithms while maintaining computational efficiency.
• To address the trade-off between speed and sensitivity inherent in existing alignment methods.
• To develop a finely grained alignment approach that preserves biological relevance in sequence comparisons.
• To extend the capabilities of prior algorithms like MASAA and MASAA Sensitive through enhanced seed selection and alignment refinement.

Proposed method

• The algorithm employs a suffix tree data structure to efficiently identify initial anchor points in sequences.
• It applies adaptive seeds that dynamically adjust based on local sequence characteristics to improve match detection.
• Finely grained perfect match seeds are used between anchors to refine alignment accuracy.
• The method integrates multiple stages of alignment refinement, focusing on local regions with high similarity.
• Sensitivity is enhanced by prioritizing high-quality seed matches while preserving computational tractability.
• The algorithm is evaluated on both synthetic sequences and the Rosetta dataset, representing diverse biological sequence lengths.

Experimental results

Research questions

• RQ1 How can heuristic sequence alignment algorithms achieve higher sensitivity without incurring excessive computational cost?
• RQ2 To what extent do adaptive and finely grained perfect match seeds improve alignment accuracy compared to fixed seed strategies?
• RQ3 Can suffix tree-based indexing effectively support high-sensitivity alignment in long sequences (up to 500k bases)?
• RQ4 How does MMSAA FG compare to prior methods like MASAA and MASAA Sensitive in terms of alignment quality and runtime?

Key findings

• MMSAA FG achieves significantly improved alignment sensitivity over previous heuristic methods, particularly in detecting subtle sequence relationships.
• The use of adaptive and finely grained perfect match seeds enhances the detection of biologically relevant matches in complex regions.
• The algorithm maintains high performance on long sequences, successfully processing datasets up to 500,000 bases in length.
• Evaluation on the Rosetta dataset confirms the algorithm's robustness and accuracy in real biological sequence alignment tasks.
• The integration of suffix trees enables efficient indexing and rapid identification of anchor points, supporting scalable alignment.
• The method demonstrates strong sensitivity gains without a prohibitive increase in computational overhead, validating its practicality.