Jack Colicchio from Sound Agriculture in the US presented at the Nanopore Community Meeting in Boston on “DNA methylation patterns explain missing heritability in maize.” This ten-minute session was part of a breakout group. Colicchio spoke about the need for more resilient crops to address pollution, climate change, and population changes. They spoke about how, traditionally, speed has been the major challenge in using predictive genomic models in plant breeding. Colicchio noted that many models only account for fifty percent of the variation. ONT long-read sequencing can provide DNA methylation from native DNA sequences. The team grew and sequenced ~200 plants and then identified epigenetic associations. The epigenetic data was then used to build a hybrid model to identify regions of the methylome strongly associated with corn cob weight. Using a machine learning framework, the model’s predictive power was improved with epigenetic data. The epipeaks were found in low-diversity regions and were often associated with the expression of nearby genes. The group is going to continue analyzing the data and confirming with phenotypic and methylation studies.
