In brief

The gut microbiome is a highly complex mix of different bacterial populations.

© 2019 A*STAR Genome Institute of Singapore

Solving the puzzle of human gut microbiomes

7 Nov 2019

A*STAR scientists have devised an algorithm for accurately assembling genomes, paving the way for in-depth analysis of microbial communities in the human gut.

Regardless of one’s standard of personal hygiene, bacteria coat every inch of our bodies and even live inside us. These microbes play important roles in maintaining health, and imbalances in their populations can result in disease.

Scientists interested in studying microbial communities, or microbiomes, often rely on a technique called metagenomics, in which bacteria are obtained from their native environment and processed for DNA sequencing. This is especially useful for studying gut microbiomes since some gut bacteria are difficult to grow in the lab.

However, metagenomic studies come with limitations. Sequencing short DNA fragments from a community of bacteria comprising hundreds of different species means that DNA fragments, or reads, need to be accurately assembled, much like the pieces of a very complex jigsaw puzzle.

“Current metagenomics assemblers only provide fragmented assemblies when there are multiple strains of the same species in the microbiome. Microbiome studies are then limited by the resolution of genetic analysis and the ability to understand microbial functions in communities harboring hundreds of bacterial species,” said Denis Bertrand, a Staff Scientist at A*STAR’s Genomic Institute of Singapore (GIS).

Together with collaborators across Singapore, including clinicians from Tan Tock Seng Hospital, and colleagues in Croatia, Bertrand sought to increase the accuracy of genome assembly in metagenomics studies using long Nanopore reads.

Assessing 197 stool samples from ongoing clinical studies, the team devised a method to analyze a majority of the samples and obtain high-quality data for long-read sequencing. By combining this data with accurate short reads, the researchers developed a hybrid assembly algorithm, OPERA-MS, which allowed them to assemble individual genomes of strains in the bacterial community from billions of DNA sequences.

“We found that OPERA-MS provides up to ten times more complete genomes compared to methods based on short-reads, and at least five times more accurate genomes than other approaches that rely on long reads,” Bertrand said. When used to analyze the gut microbiomes of 28 antibiotic-treated patients, OPERA-MS facilitated the discovery of gene combinations responsible for resistance to several antibiotic classes.

“These assemblies serve as valuable references for studying the evolution of antibiotic-resistant microbes in the gut. We can now distinguish between antibiotic-resistant strains and those that are benign residents, allowing us to track the spread of both infectious and beneficial bacteria in their natural ecosystem,” Niranjan Nagarajan, Associate Director and Senior Group Leader at GIS, explained.

As part of the integrated Omics research program at A*STAR, the team plans to use OPERA-MS to generate an Asian gut bacterial reference genome and identify microbiome variations between and within ethnic groups. In collaboration with Hong Kong-based Civet Bioscience, OPERA-MS will also be used to monitor the microbiome of patients who have undergone fecal microbiota transplantation.

The A*STAR-affiliated researchers contributing to this research are from the Genomic Institute of Singapore (GIS).

Want to stay up to date with breakthroughs from A*STAR? Follow us on Twitter and LinkedIn!


Bertrand, D., Shaw, J., Kalathiyappan, M., Ng, A. H. Q., Kumar, M. S. et al. Hybrid metagenomic assembly enables high-resolution analysis of resistance determinants and mobile elements in human microbiomes. Nature Biotechnology 37, 937-944 (2019) | article

About the Researcher

Denis Bertrand obtained his PhD degree at the University Montpellier II, France, where he worked on the development of computational methods to study the evolution of tandem duplicated genes. He completed his postdoctoral training at the University of Montreal, Canada. Bertrand is currently a Staff Scientist at the Genome Institute of Singapore (GIS), working in the lab of Niranjan Nagarajan, Associate Director and Senior Group Leader at GIS. His research focuses on the development of tools applied to genomic data, including genome assembly methods for single and metagenomes, and integrative approaches for the inference of cancer driver genes.

This article was made for A*STAR Research by Wildtype Media Group