Sepia, a taxonomy oriented read classifier in Rust

Area of Science:

• Bioinformatics
• Computational Biology
• Genomics

Background:

• Accurate taxonomic classification of biological sequences is crucial for understanding microbial communities and evolutionary relationships.
• Existing read classifiers may face challenges with speed, accuracy, and handling diverse or inconsistent taxonomic databases.

Purpose of the Study:

• To introduce Sepia, a novel, high-performance read classifier designed for speed and accuracy.
• To provide a flexible tool capable of managing multiple taxonomic frameworks and identifying database inconsistencies.

Main Methods:

• Sepia is implemented in the Rust programming language for optimal performance.
• The classifier allows dynamic switching between various taxonomic databases.
• It incorporates algorithms for detecting inconsistencies within taxonomic hierarchies.
• Similarity estimation between query sequences and reference databases is a core functionality.

Main Results:

• Sepia demonstrates fast and accurate performance in read classification tasks.
• The tool effectively identifies and flags inconsistencies present in taxonomic databases.
• It provides quantitative measures of similarity between biological samples and reference data.

Conclusions:

• Sepia offers a robust and efficient solution for large-scale genomic data analysis.
• Its ability to handle taxonomic variations and inconsistencies enhances the reliability of biological sequence classification.
• The tool is valuable for researchers in genomics, metagenomics, and evolutionary biology.