GPMeta

GPMeta provides ultrarapid pathogen identification from metagenomic next-generation sequencing (mNGS) data by using GPU acceleration and statistical clustering to improve clinical microbiological diagnostics.

Key Features:

• GPU Acceleration: Uses Graphics Processing Units (GPUs) to accelerate computational processes for pathogen identification, reducing analysis time relative to CPU-based methods.
• High Accuracy and Speed: Demonstrates higher accuracy and greater speed than Bowtie2, Bwa, Kraken2, and Centrifuge on mock microbial community datasets and clinical metagenomic sequencing data.
• GPMetaC Clustering Algorithm: Implements the GPMetaC statistical clustering algorithm to cluster and rescore ambiguous alignments, improving discrimination of highly homologous microbial genomes (average nucleotide identity >95%) and enhancing precision and recall.

Scientific Applications:

• Infectious disease diagnostics: Facilitates pathogen identification from mNGS data for clinical microbiological testing and diagnostic workflows.
• Infection control and patient care: Supports timely decision-making in infection control and patient care by reducing turnaround time for pathogen classification.

Methodology:

Performs GPU-accelerated sequence alignment combined with advanced statistical models, specifically the GPMetaC clustering algorithm that clusters and rescoring ambiguous alignments to discriminate genomes with average nucleotide identity >95%.