A systematic comparison of human mitochondrial genome assembly tools

Area of Science:

• Genomics
• Bioinformatics
• Cell Biology

Background:

• Mitochondria generate cellular energy and possess their own genome, crucial for biotechnology and phylogenetics.
• Independent mitochondrial genomes encode essential genes and are targets for various assembly tools.
• Existing tools for mitochondrial genome assembly often use whole-genome sequencing data.

Purpose of the Study:

• To systematically compare available tools for assembling human mitochondrial genomes using short-read sequencing data.
• To identify optimal assemblers for diverse research scales, from small projects to national initiatives.

Main Methods:

• Evaluated mitochondrial genome assemblers using simulated and human whole-genome sequencing (WGS) datasets.
• Assessed performance based on execution time, computational memory usage, and accuracy (SNP detection).
• Tested tools across various sequencing depths and computational thread settings.

Main Results:

• MitoFlex and IOGA showed significant differences in execution time and memory usage on simulated data.
• GetOrganelle and MitoFlex demonstrated superior SNP capture on human WGS data (mean F1-score 0.919 at 10X depth).
• MToolBox and NOVOPlasty exhibited consistent performance across sequencing depths (mean F1 scores 0.897 and 0.890).

Conclusions:

• MToolBox is recommended for overall performance and assembly quality across sequencing data types.
• NOVOPlasty is a practical choice for large-scale projects with limited computational resources due to its speed and low memory footprint.
• Future development of mitochondrial genome assemblers should address the growing use of long-read sequencing technologies.