VeChat

VeChat corrects errors in long-read sequencing data using variation graphs to reduce sequencing errors while preserving true genetic variants for haplotype-aware assembly and variant analysis.

Key Features:

• Graph-Based Reference System: Leverages variation graphs as pangenome reference structures to avoid consensus-sequence biases that can mask true variants.
• Error Reduction: Reduces errors by 4–15× for Pacific Biosciences reads and 1–10× for Oxford Nanopore Technologies reads compared to state-of-the-art approaches.
• Haplotype Awareness: Preserves and improves detection of low-frequency haplotypes in mixed samples to support accurate haplotype-aware assemblies.
• Pre-assembly Correction: Operates prior to long-read assembly to enhance downstream assembly accuracy and variant representation.

Scientific Applications:

• Variant Calling: Improves accuracy of variant detection by reducing sequencing errors while retaining true genetic variation.
• Genome Assembly: Enhances long-read assembly quality and haplotype resolution when used as a pre-assembly error-correction step.
• Population Genetics: Enables more reliable analysis of genetic diversity by preserving low-frequency variants across samples.
• Mixed-sample Analysis: Facilitates accurate representation of haplotypes in mixed or metagenomic samples by avoiding consensus-induced biases.

Methodology:

Leverages variation graphs (pangenome reference structures) to perform error correction of long reads prior to long-read assembly.