Unraveling the role of the microbiome in human health and environmental samples is an emerging priority in scientific study. But despite the best advances in sequencing technology, identifying the bacteria, fungi, and other organisms present in complex samples remains a huge challenge.
Metagenomic shotgun sequencing can read chromosomes, plasmids, and bacteriophages, and comparison to reference genome sequences can be used to place them into putative taxa and species bins, but these methods fail to sufficiently distinguish between genomes that are very similar.
A team of scientists from the Icahn School of Medicine at Mount Sinai, Sema4, and other institutions has come up with a novel solution: a computational method that uses PacBio long-read sequencing of metagenomic DNA to identify methylated motifs and create an epigenetic barcode that enables more precise microbiome analysis.
The process takes advantage of methyl groups which are added to nucleotides in bacteria and archaea in a highly sequence-specific manner, and these motifs often differ among species and strains.
The team took advantage of inter-pulse duration values that represent the time it takes a DNA polymerase to translocate from one nucleotide to the next during SMRT Sequencing. This measure can distinguish between methylated and non-methylated bases. They calculated methylation scores across motifs of several bacterial samples and murine fecal samples and created methylation profiles, which were used alongside sequence composition features to assemble contigs into species- and strain-level bins.
In a paper published in Nature Biotechnology, senior author Gang Fang describes how the method was also able to link mobile genetic elements, including antibiotic resistance-encoding plasmids, to their host species in a real microbiome sample.
Although their sequence coverages and composition profiles often differ, plasmid and chromosomal DNA of the bacterial host are methylated by the same set of methyltransferases, resulting in matching methylation profiles, the authors note.
“The biomedical community has long needed a microbiome analysis method capable of resolving individual species and strains with high resolution,” Fang said in statement.
The method could ultimately prove useful in both research and clinical settings, since it allows for linking mobile genetic elements to their bacterial hosts. This information makes it possible for scientists to more accurately predict virulence and antibiotic resistance of individual bacterial species and strains, among other important traits.