Impact of sorting in DNA sequence compression

• Main Author: Fonseca, Tiago Rafael Soares da
• Publication Date: 2023
• Format: Master thesis
• Language: eng
• Source: Repositórios Científicos de Acesso Aberto de Portugal (RCAAP)
• Download full: http://hdl.handle.net/10773/41955
• topic: Lossless data compression; DNA sequences; Genome; Sorting; Efficiency

The increase in the production of genomic data led to a growing need to find efficient methods that could store and analyze this type of data. Due to its redundant structure and the storage space needed, data compression is seen as an essential strategy to deal with this kind of problem. There are various tools available, from those that were created with a more general use intent to those that were created to deal specifically with this type of data. There’s however still space to test new tools who can make this process more efficient. There’s the case of the tool presented here, FASTA ANALYSIS, which allows the sorting of FASTA sequences (a genomics sequence representation type) according to different criteria, taking advantage of putting closer similar blocks. The benchmark performed had as an objective to compare the impact of sorting two distinct file types. On one hand, there were generated nearly-synthetic sequences with the AlcoR toolkit, who permits to build sequences from specific input data. On the other hand, there were used groups of specific genomes and collections of genomes organized according to their type (bacteria, viral, etc). The results here presented show variable gains depending on the file being tested, but they’re more pronounced when the files are sorted by the absolute number of bases. Furthermore, it was also possible to notice that the file types that showed higher percentage gain, among those tested, were those who combine different types of datasets. FASTA ANALYSIS is a tool which allows the improvement of compression efficiency of FASTA sequences, by sorting the sequences present on Multi-FASTA files, according to different criteria. FASTA ANALYSIS is distributed under GPLv3 and it’s available for free download at https://github.com/tiagof1993/FASTA-ANALYSIS.