Multi-sample graph-based framework for comparative microbiomes genomics

This technology is a multi-sample, sequence-graph-based framework that improves the accuracy and resolution of metagenomic analysis across microbiomes.

Unmet Need: Accurate, scalable microbiomes genomic analysis that facilitates detection of strain-level differences

Current methods for analyzing metagenomic sequencing data from microbiomes struggle to deliver accurate results at a high resolution that parallels analysis of microbial isolates without high computational costs and complexity. These limitations prevent a deep analysis of genomic differences in microbiome, reduce scalability and hinder the integration of multi-sample datasets needed to capture microbial diversity and dynamics. As a result, researchers face barriers in extracting meaningful biological insights from microbiome data. There is a need for efficient, high-resolution, and high-accuracy approaches that enable scalable analysis across complex microbiomes.

The Technology: High-resolution and computationally efficient framework for microbiome genomics

This technology is a multi-sample, sequence-graph-based framework that enables accurate comparisons of genomic data across microbiomes with low computation cost, paralleling comparative genomics without the need for isolate sequencing. The analysis software achieves this by integrating hybrid coassembly, homology-based graph merging, and graph-optimization algorithm into a unified platform to improve accuracy and scalability of metagenomic analysis.

This technology has been validated in multiple settings, including colorectal cancer, preterm birth, and Vancomycin-resistant Enterococcus.

Applications:

  • Software for comparative metagenomic analysis
  • Genome language modeling for artificial intelligence applications
  • Biosecurity screening for harmful microorganisms

Advantages:

  • Computationally efficient
  • Improves accuracy of sequence and variant detection
  • Optimized for comparative genomic analysis

Lead Inventor:

Tal Korem, Ph.D.

Patent Information:

Patent Pending

Related Publications:

Tech Ventures Reference:

Quick Facts:
Inventors
Izaak ColemanTal Korem
Manager
Joan Martinez
Departments
Cellular, Molecular and Biomedical StudiesSystems Biology
Divisions
Columbia University Medical Center
Reference Number
CU26050
Release Date
2026-05-21