The German Conference on Bioinformatics (GCB) is an annual, international conference devoted to all areas of bioinformatics and meant as a platform for the whole bioinformatics community.
Computomics is a highly innovative bioinformatics company focussing on agricultural challenges in plant breeding. Intense pressure is put on farmers, growers and agricultural companies to produce new plant varieties quickly while meeting the global needs of sustainability. Computomics has developed ⨉SeedScore®, a disruptive machine learning technology which transforms traditional plant breeding processes. xSeedScore supports traditional plant breeding methods and integrates additional variables such as information on the genotype, the cultivation environment, climate, soil microbiome, and the used field management to identify and predict the best-performing plants for any specific location. By applying Computomics' machine learning technology ⨉SeedScore® to a plant breeding program can help move the program way beyond current limitations. It increases the probability of success by identifying up to 10x more candidates to escalate into the commercial pipeline, and predicts best performers, both for today and future climates.
On Tuesday, 6 September 2022, Patrizia Ricca will present a poster on how we can gain insight into biological mechanisms using machine learning.
Title: Peek into the black box: How we can gain insight into biological mechanisms using machine learning
When: Tuesday 6 September 2022, 17:30 - 19:30
Patrizia Ricca, Scientific Product Manager at Computomics
Plant breeding must be accelerated to supply new varieties for a growing population and a rapidly changing climate. New breeding technologies like gene editing and genomic prediction help bring about this acceleration, but are often used independently without sharing useful existing knowledge. Here, we present a method for discovering both new gene editing targets and higher-accuracy predictions. By using interpretable machine learning models specifically developed for genomic data, complex genetic mechanisms can be rapidly understood and visualized. Multigenic traits show up in the visualization of feature importance and positional genomic importance. We apply this method to a dataset derived from a shelf-life experiments for 200 Capsicum varieties. Genotypes, manual scoring and plant image data are correlated to train a regression machine learning algorithm that identifies an ethylene-linked gene cluster responsible for shelf life and plant senescence. New breeding technologies require these kinds of insights into biological regulation to identify new editing targets quickly and reliably.
Meet and connect with Patrizia in Halle at GCB. In case of questions regarding the poster or to discuss your plant breeding challenges please feel free to contact Patrizia directly.