Developing the Apollo software for high-throughput annotation of multiple genomes

开发用于多个基因组高通量注释的 Apollo 软件

• 批准号: 10736567
• 负责人: Ian H Holmes
• 金额: $ 46.17万
• 依托单位: UNIVERSITY OF CALIFORNIA BERKELEY
• 依托单位国家: 美国
• 财政年份: 2007
• 资助国家: 美国
• 起止时间: 2007-08-01 至 2027-08-31
• 项目状态: 未结题
• 来源: https://reporter.nih.gov/project-details/10736567
• 关键词: Artificial Intelligence; Computer software; Data; Generations; Genes; Genome; Genomic DNA; Individual; Internet; Location; Machine Learning; Maintenance; Manuals; Organism; Phylogenetic Analysis; Research Personnel; Source; Synteny; System; Training; Variant; computerized tools; crowdsourcing; empowerment; gene function; genome annotation; genome sequencing; genomic data; improved; interest; outreach; tool; usability; volunteer

Genome sequencing is becoming cheaper and more powerful. However, a bottleneck to scientific progress using these data is the generation of high-quality genome annotations, describing the location and function of the genes in genomic DNA. In the past, these annotations were generated exclusively by professional biocurators, and these biocurators continue to be a critical source of expertise. However, the number of such biocurators is small compared to the amount of genomic DNA that is available. In recent years, professional biocuration has been supplemented by a growing volunteer force of interested biologists, each generally interested in one or two genes of interest. Although genome annotation is often done in bulk using computational tools, there is currently no substitute for a final pass of manual annotation. This is most commonly done using our Apollo software, which allows for live simultaneous collaborative annotation over the web. Our proposal here is to improve Apollo so as to empower both professional biocurators and crowdsourcing volunteers. We will empower professional biocurators by giving them the power tools they need to simultaneously annotate multiple genomes (by exploiting synteny), annotate variants, and annotate the function of genes. We will empower the crowdsourcing volunteers by lowering barriers to entry, making Apollo more usable. For all users we will train machine learning systems to automatically detect common annotation errors and suggest improvements. We will support Apollo users with maintenance, bugfixing, and various feature requests, as well as extensive outreach including outreach to developers.

基因组测序正变得越来越便宜和强大。然而，瓶颈来了