This AI Learns Biology To Predict Perfect Crops From Just DNA скачать в хорошем качестве

This AI Learns Biology To Predict Perfect Crops From Just DNA

How can we create better crops to feed a growing world? Scientists are turning to artificial intelligence. Traditionally, predicting a plant's traits like yield or drought resistance from its DNA has been difficult and only moderately accurate. But a new approach is changing the game. Researchers have developed a 'Biology-Informed Neural Network', or BINN. Think of it like giving an AI a biology textbook to study before an exam. During its training, the AI is fed not just DNA data, but also complex information about gene activity and other molecular processes. This teaches it the deep, often non-linear, rules that connect genes to traits. The best part? Once it's trained, it only needs a simple DNA sample to make its predictions. This method has shown incredible results. In tests with maize, it improved prediction accuracy by up to 56. In a computer-simulated test where the true biology was known, it slashed prediction errors by 75 and correctly identified the most critical biological pathway. This technology could dramatically accelerate crop breeding, allowing scientists to design more resilient and productive plants to help ensure global food security. Cited paper: K. Kontolati et al. (2025). Biology-informed neural networks learn nonlinear representations from omics data to improve genomic prediction and interpretability. arXiv:2510.14970v1. http://arxiv.org/abs/2510.14970v1 Images shown are page renders from the paper PDF for commentary/education.