Crop-GPA: an integrated platform of crop gene-phenotype associations

• Published in: NPJ systems biology and applications Vol. 10; no. 1; pp. 15 - 10
• Main Authors: Gao, Yujia, Zhou, Qian, Luo, Jiaxin, Xia, Chuan, Zhang, Youhua, Yue, Zhenyu
• Format: Journal Article
• Language: English
• Published: London Nature Publishing Group UK 12.02.2024; Nature Publishing Group; Nature Portfolio
• Subjects: 631/114; 631/449; Artificial intelligence; Automation; Bioinformatics; Biology; Biomedical and Life Sciences; Computational Biology/Bioinformatics; Computer Appl. in Life Sciences; Crop diseases; Crops; Deep learning; Genes; Genetically altered foods; Genomes; Genomics; Genotype & phenotype; Life Sciences; Machine learning; Phenotypes; Plant breeding; Systems Biology
• ISSN: 2056-7189; 2056-7189
• DOI: 10.1038/s41540-024-00343-7

With the increasing availability of large-scale biology data in crop plants, there is an urgent demand for a versatile platform that fully mines and utilizes the data for modern molecular breeding. We present Crop-GPA ( https://crop-gpa.aielab.net ), a comprehensive and functional open-source platform for crop gene-phenotype association data. The current Crop-GPA provides well-curated information on genes, phenotypes, and their associations (GPAs) to researchers through an intuitive interface, dynamic graphical visualizations, and efficient online tools. Two computational tools, GPA-BERT and GPA-GCN, are specifically developed and integrated into Crop-GPA, facilitating the automatic extraction of gene-phenotype associations from bio-crop literature and predicting unknown relations based on known associations. Through usage examples, we demonstrate how our platform enables the exploration of complex correlations between genes and phenotypes in crop plants. In summary, Crop-GPA serves as a valuable multi-functional resource, empowering the crop research community to gain deeper insights into the biological mechanisms of interest.