Split-alignment of genomes finds orthologies more accurately

Area of Science:

• Genomics
• Bioinformatics
• Comparative Genomics

Background:

• Accurate genome alignment is crucial for comparative genomics.
• Previous methods often mask repeats or assume long, un-rearranged blocks, limiting accuracy.
• Existing animal genome alignments contain errors and misalignments.

Purpose of the Study:

• To develop a novel pair-wise genome alignment method.
• To improve the accuracy of genome alignments by optimizing local alignments.
• To identify non-orthologous regions and uncover evolutionary features missed by previous methods.

Main Methods:

• A new pair-wise genome alignment approach based on optimal local alignments.
• Incorporation of a statistical model to quantify alignment ambiguity.
• Avoidance of repeat masking to preserve alignment information.

Main Results:

• The new method aligns thousands of locations differently compared to previous methods.
• Alignments show significantly higher similarity, indicating improved accuracy.
• Identified numerous potentially non-orthologous regions in animal genomes.

Conclusions:

• The novel alignment method offers higher accuracy and resolves issues with previous approaches.
• The findings suggest widespread non-orthology in existing animal genome alignments.
• This method facilitates the discovery of evolutionary features like fast-evolving elements and micro-rearrangements.