Researchers from U of T, the Ontario Institute for Cancer Research, and Johns Hopkins University have published a study in Nature Methods showcasing their new method for detecting cytosine methylation, an epigenetic mark on DNA.
Our genetic makeup, DNA, is comprised of four essential building blocks. These building blocks are known as nucleotides called adenine, thymine, guanine, and
cytosine. The addition of a methyl group to any of these nucleotides is called methylation. This study focused on the methylation of cytosine, which has been shown to turn on and off genes that play a role in disease.
Since cytosine methylation has been linked to diseases such as cancer, researchers are trying to understand and map methylation patterns in humans. Existing methylation mapping techniques require a large input of DNA and harsh chemical treatment, which causes the DNA to break into fragments.
The Oxford Nanopore Technologies MinION is a commercially available sequencing technology that measures an electrical current and how that current is disrupted when a DNA fragment moves through a pore.
Dr. Jared Simpson, an Assistant Professor at U of T and first author on the study, said that he “became interested in nanopore sequencing because it overcomes a lot of the issues of conventional short read sequencers — the MinION is portable and it can read incredibly long fragments of DNA.”
After developing algorithms to interpret and map DNA, he decided to dive deeper into understanding and unraveling the electrical signals within the nanopore data.
In a Nature Methods paper co-authored by Simpson, the researchers examined the electrical signals from the MinION, a sequencer the size of a USB drive. The researchers used methylated and unmethylated Escherichia coli data to create and train a computational algorithm to distinguish signal patterns from cytosines that were methylated from those that were not.
After creating this model, the authors tested their algorithm on human cell lines and positive and negative controls to confirm that the algorithm was able to differentiate between methylated and unmethylated cytosines. The researchers hope to extend this analysis to human tumours in the future.
To promote research in this field, the analysis pipeline developed for the detection of DNA methylation via the MinION is completely open-source and available freely online.
The study represents one of the first attempts to use the electrical output from the MinION sequencer, without any treatment of the DNA, to detect cytosine methylation. In the same issue of Nature Methods, a study led by Benedict Paten demonstrated a similar approach of detecting different types of DNA methylation using the MinION, further supporting the novelty and utility of this type of research.
While cytosine methylation is the most common form of methylation in humans and has been well characterized in cancer, there are many other forms of methylation. Simpson hopes to begin acquiring training data for other methylation marks.
According to Simpson, this combined method of mapping both the basic building blocks of DNA and cytosine methylation is only a stepping stone. In the future, he hopes that nanopore sequencing will be used as a single platform to examine many types of genetic variation.
“The signals measured by nanopore sequencers [are] a very rich source of information,” Simpson said. “We showed that this extra information is accessible if you carefully build models of the signal data.”