Improving Gene Annotation with Machine Learning
Advancements in genome sequencing have greatly enhanced the study of organisms, but understanding genomic variation remains challenging, especially in genome annotation, which models gene transcription and translation. 🧬
We refined a machine learning model called the Genomic Pre-trained Network (GPN), significantly improving gene annotation in plant genomes. By utilizing ribosome profiling data, we accurately identified gene initiation and stop sites, achieving a 92% prediction accuracy. 📈
Note: The GPN model demonstrates the power of integrating advanced machine learning with genomic data, improving the precision of genome annotation and accelerating research in plant genomics.
https://github.com/maize-genetics/gene_modeling