SeqDeχ: A Sequence Deconvolution Tool for Genome Separation of Endosymbionts From Mixed Sequencing Samples

Area of Science:

• Genomics
• Bioinformatics
• Computational Biology

Background:

• Next-generation sequencing (NGS) has reduced genome sequencing costs, enabling the analysis of complex symbiotic consortia.
• Analyzing mixed DNA samples from symbiotic organisms requires specialized bioinformatics tools for accurate data deconvolution.

Purpose of the Study:

• To introduce SeqDex, a novel, automated bioinformatics tool for identifying specific organism contigs from mixed DNA sequencing data.
• To provide a user-friendly and customizable solution for deconvolving host-symbiont datasets.

Main Methods:

• SeqDex utilizes a machine learning approach, combining partial taxonomic affiliations and compositional analysis to predict contig origins.
• The tool processes preliminary assemblies from mixed DNA samples to identify sequences belonging to a target organism.
• Performance was evaluated by comparing SeqDex against state-of-the-art metagenomic binning and host-symbiont deconvolution methods on three case studies.

Main Results:

• SeqDex demonstrated strong performance in deconvolving host-symbiont datasets across multiple study cases.
• The tool effectively automates the identification of contigs from specific organisms within mixed samples.
• Compared to existing methods, SeqDex offers improved ease of use and customization potential.

Conclusions:

• SeqDex is an effective and efficient tool for the rapid identification of endosymbiont sequences from complex symbiotic consortia.
• The automated, machine learning-based approach of SeqDex addresses the need for specialized bioinformatics tools in symbiotic research.
• SeqDex offers a valuable resource for researchers studying host-symbiont interactions and microbial communities.