November 21, 2019
International research group unlocks the promise of nanopore native RNA sequencing
Studying RNA may offer new answers to cancer – and the tools to read RNA directly are now in our hands.
An international research consortium, led in part by Dr. Jared Simpson at OICR, has developed new laboratory protocols and a suite of software tools that will allow the research community to exploit the promise of direct RNA sequencing.
These techniques, published recently in Nature Methods, represent the first large-scale exploration of human RNA using nanopore sequencers – the advanced handheld sequencing devices that can read long strands of RNA.
“Unlike traditional sequencing devices that read copies of RNA strands that are cut into little pieces, nanopore sequencing allows us to study long strands of RNA directly without losing important information in the copying and cutting process,” says Paul Tang, Computational Biologist at OICR and co-first author of the publication. “Our methods combine the power of reading RNA directly with the power of long-read sequencing, enabling an entirely novel way to study cancer biology.”
In collaboration with researchers at Johns Hopkins University and the University of California Santa Cruz, Tang and Simpson developed the software methods that could decode the output data from a nanopore sequencer. Their methods used a machine learning technique, called a Hidden Markov Model, to determine the letters of code within an RNA strand.
“With these methods, we’ve shown that you can leverage nanopore RNA sequencing to gain a lot of valuable information that we couldn’t have otherwise,” Tang says. “We’re very happy to see this work published because we are enabling others to study a new aspect of cancer biology and we look forward to the research discoveries to come.”
These new methods have been integrated into Simpson’s already-popular nanopolish software suite which is routinely used by the nanopore community around the world.
September 18, 2019
Dr. Ina Anreiter joins OICR as a Schmidt Fellow, bringing her background in behavioural genetics to bioinformatics
While writing her doctoral thesis, Dr. Ina Anreiter realized that there was a missing piece to her research. What she didn’t realize was that this missing piece would lead her into a prestigious postdoctoral fellowship in an entirely new scientific discipline. For decades, scientists have known that RNA – often referred to as DNA’s cousin – undergoes chemical modifications before running its course. These modifications, like RNA methylation, have an important effect in cancer cells, but without the tools to study RNA modifications, progress in this field had stalled for many years.
Recently, the study of these modifications – also known as the field of “epitranscriptomics” – has garnered new attention as the research community develops new methods to study RNA. These methods, Anreiter says, still rely on common chemistry lab techniques and cumbersome procedures that make studying RNA methylation difficult, especially in application to diseases like cancer.
“I found myself in need of a tool,” says Anreiter. “I needed a way to easily analyze RNA methylation across large datasets and found that nothing existed – well, nothing existed yet.”
From fruit flies to machine learning
Anreiter’s doctoral research focused on the behaviour of fruit flies, specifically how inherited characteristics and environmental factors influence their feeding patterns. While searching for a way to study RNA methylation, her background led her to a unique idea.
Anreiter knew of nanopore sequencing – a relatively new type of sequencing technology that could decode DNA and RNA as it passes through a tiny channel. By directly reading a strand of RNA, Anreiter says, nanopore sequencing has the potential to revolutionize how we study RNA modifications. To this day, however, there are no algorithms or tools that can accurately find RNA methylation patterns in the output data of a nanopore sequencer.
Anreiter had also heard of Dr. Jared Simpson’s breakthrough methods for detecting DNA methylation using nanopore sequencing. His computational methods allowed the nanopore community to sequence the entire – highly-methylated – human genome in 2017, and since, he has been working in part to study RNA modifications, like RNA methylation, using nanopore sequencing.
Anreiter pitched her idea to Simpson.
“RNA methylation occurs in normal fruit flies, but not in a certain type of mutant fly,” says Anreiter. “I had a crazy idea that we could sequence both of these types, and use the datasets to develop a machine learning algorithm that could find RNA methylation on its own.”
The potential of her idea would win her the prestigious Schmidt Science Fellowship and a $100,000 USD stipend to work with Simpson for a year.
From machine learning to cancer patients
Anreiter recently began her year-long postdoctoral fellowship in the Simpson Lab at OICR where she is working alongside a team of computational biologists to turn her idea into an algorithm. She is cross-appointed with the University of Toronto’s Department of Computer Science.
“At this point, we’re working on a preliminary dataset, but I’ve already learned so much. The team has been very welcoming and supportive and we’re working together to make better tools to understand diseases.”
The Schmidt Fellowship, which was co-founded by the former CEO of Google, is awarded to exceptional, early-career researchers making a “pivot” in their work. Anreiter saw the fellowship as an opportunity to immerse herself in a completely new field.
“If we can develop this tool, it would allow us to study human diseases in a new way,” Anreiter says. “When we look at a problem in a new way, we don’t know what solutions we’ll find, but this angle could lead us to new cures.”