CSA-MEM: Enhancing Circular DNA Multiple Alignment Through Text Indexing Algorithms

In the realm of Bioinformatics, the comparison of DNA sequences is essential for tasks such as phylogenetic identification, comparative genomics, and genome reconstruction. Methods for estimating sequence similarity have been successfully applied in this field. The application of these methods to circular genomic structures, common in nature, poses additional computational hurdles. In the advancing field of metagenomics, innovative circular DNA alignment algorithms are vital for accurately understanding circular genome complexities. Aligning circular DNA, more intricate than linear sequences, demands heightened algorithms due to circularity, escalating computation requirements and runtime. This paper proposes CSA-MEM, an efficient text indexing algorithm to identify the most informative region to rotate and cut circular genomes, thus improving alignment accuracy. The algorithm uses a circular variation of the FM-Index and identifies the longest chain of non-repeated maximal subsequences common to a set of circular genomes, enabling the most adequate rotation and linearisation for multiple alignment. The effectiveness of the approach was validated in five sets of mitochondrial, viral and bacterial DNA. The results show that CSA-MEM significantly improves the efficiency of multiple sequence alignment, consistently achieving top scores compared to other state-of-the-art methods. This tool enables more realistic phylogenetic comparisons between species, facilitates large metagenomic data processing, and opens up new possibilities in comparative genomics.