By: Louisa Gaylord

Korena Mafune is a 2021 Washington Research Foundation postdoctoral fellow in the University of Washington’s Department of Civil and Environmental Engineering, where she studies the symbiotic interactions among plants, fungi, and bacteria. She received her PhD from UW’s School of Environmental and Forest Sciences, where she studied the root-associated fungal communities of old-growth bigleaf maple trees in Washington’s Queets and Hoh temperate rainforests. These trees accumulate layers of leaves and other organic matter on their canopy branches, which decay over time and produce a thick mat of organic soil high above the forest floor. Bigleaf maple trees have the capability to grow extensive adventitious rooting networks into these canopy soils which associate with fungal communities that thrive in the damp Pacific Northwest forests; some of these fungal associates attach to the roots and expand their fungal network outwards to aid the tree in taking up plant nutrients like nitrogen and phosphorus which the trees rooting system may not be able to reach on its own.

During her PhD, Mafune explored the structures and community diversity of root associated fungi, soil micro-climate, and nutrient dynamics between the two soil environments (both in the forest floor and canopy soils) of several bigleaf maple trees. Her goal was to determine whether there was a noticeable difference between the amount of available plant nutrients that the systems produced, how this is tied to soil temperature and moisture, and how fungal mutualists in the canopy soils may be accessing these nutrients to benefit the host tree. To better understand these environments, she sampled roots seasonally for over a year. She collected a total of 1,200 soil and root samples from two different rainforest sites in Western Washington and stained them for microscopy to observe their fungal structures. Mafune then continued to sample roots and soils for DNA and nutrient analyses, and used data loggers to collect data on the soil’s temperature and moisture. Back in the lab, she isolated the tiny roots of the bigleaf maples that associate with mutualistic fungi and processed them through a DNA extraction protocol. To make sure DNA was successfully extracted from the root material, she used a Qubit fluorometer to measure the amount of DNA and then prepared the DNA for the next step: polymerase chain reaction (PCR) with specific primers to amplify the fungal DNA from the other DNA that were present in the samples.

The purified fungal DNA was barcoded and further prepared for sequencing on a MinION nanopore sequencer. At the time she was conducting this research, Korena Mafune wanted to use the MinION for its ability to provide longer sequence reads that could give more detail on fungal diversity. However, no one had ever tried to use MinION nanopore sequencer for this type of community analysis before, so she needed to benchmark the methodology to make sure it worked. Mafune created a community of known fungal species to see if the MinION could correctly determine the species; the sequencer was not only able to identify mutualistic root-associated fungi that benefit the tree’s overall health, but also other fungi that are often understudied. However, there wasn’t an established approach or bioinformatics pipeline to successfully do so since the MinION was relatively new at that time, so Mafune and one of her mentors – Bruce Godfrey, PhD – benchmarked a methodology and bioinformatics pipeline and started processing her field samples. “At this point we had very large files [of fungal DNA] that were barcoded by sample,” she explained. “I collected a root sample from the canopy of one tree and floor of the same tree, and each DNA sample has its own barcode so I could assign IDs to each sample… We gained a lot of fundamental information about root-associated fungi and nutrient dynamics in canopy soils, demonstrating these unique environments should not be overlooked, especially in a changing climate.”

Although the MinION nanopore sequencer was able to provide species-level inferences on root-associated fungal communities, Mafune also sequenced the samples on an illumina MiSeq to provide supplemental sequencing data. Illumina is a well established high-throughput sequencing technology, but it doesn’t have the ability to provide species-level inferences like the MinION could. As a brand new sequencer, MinION didn’t have an established pipeline of previous use. “The MinION pipeline is cheaper to use and portable, so [this type of] fungal community analysis could really benefit smaller scientific communities or even farming communities,” Mafune said. “We started benchmarking this about five or six years ago, and we have seen it advance and improve so much over the years, the output and results are only getting stronger.”

Korena Mafune and Bruce Godfrey then utilized a series of Python scripts to perform a variety of tasks: pulling out DNA sequences by their barcode, trimming the identifier name, and putting them into respective operational taxonomic unit (OTU) clusters based on similarity, as well as creating a consensus sequence that could be manipulated in the database. Bruce Godfrey, who is a research scientist as well as a colleague of Mafune’s, created several of the scripts that were used. The OTU dataset was imported to Rstudio, along with the MinION and illumina MiSeq datasets, to analyze the diversity and implement other statistical tests that determined if the nutrient and microclimate dynamics had an impact on root-associated fungal communities.

“I can answer questions like ‘Are fungal communities significantly different between the canopy and forest floor rooting networks?’ and ‘Do these communities shift seasonally?’ [but] there is still a lot to be studied!” said Korena Mafune. This new MinION data sequencer pipeline can benefit researchers in a number of different fields, like identifying pathogens in agricultural plants.

Data science supports  visualization and analysis of huge datasets that incorporate any number of factors across space and time. The growing need to analyze larger and more complex amounts of data means that data science can be utilized just about anywhere, both within STEM fields and beyond. “I didn’t know I’d become somebody who processes so much data and does large-scale analyses,” Mafune said. “Then I realized how important data was in the field… I wanted to collect everything I could, but first I had to know what questions I wanted to pursue.”

If people have no previous experience using data science tools, the first classes can be overwhelming. The eScience Institute regularly offers a Software Carpentries workshop that focuses on software tools to make researchers more effective, including program design, version control, data management, and task automation. Participants learn how to automate research tasks, automatically track their research over time, and use programming in Python to accelerate and make their research more reproducible. Sign up here to get notified when registration opens for the next workshop.