Xander

Xander assembles targeted protein-coding genes from metagenomic sequencing reads by integrating de Bruijn graphs with protein profile hidden Markov models (HMMs) to improve assembly accuracy and gene annotation.

Key Features:

• Integrated Graph Structure: Merges de Bruijn graph representations of sequence k-mers with protein profile HMM information to form a combined weighted assembly graph that guides targeted assembly.
• Concomitant Assembly and Annotation: Performs assembly and annotation simultaneously by incorporating HMM information into the assembly graph.
• Enhanced Gene Quality: Targets specific protein-coding genes to produce longer and higher-quality full-length or near-full-length gene sequences compared to bulk metagenome assembly methods and other gene-targeted assemblers, demonstrated on Human Microbiome Project (HMP) Illumina data and rhizosphere soil metagenomic datasets from diverse crops.
• Flexibility in Gene Targeting: Accepts customized protein profile HMMs tailored to specific target genes to improve precision of gene recovery and annotation.

Scientific Applications:

• Targeted gene recovery in metagenomes: Enables recovery of full-length or near-full-length functional protein-coding genes from complex metagenomic samples.
• Microbial functional profiling across environments: Facilitates characterization of microbial functional genes in datasets such as Human Microbiome Project (HMP) Illumina sequences and rhizosphere soil metagenomes from various crops.

Methodology:

Integrates de Bruijn graphs and protein profile HMMs to create a combined weighted assembly graph and performs concomitant assembly and annotation of targeted protein-coding genes.