Prediction of protein subcellular localization for prokaryotes, using experimentally-based criteria

在农业和医学领域，人们越来越感兴趣的是通过计算预测细菌蛋白质的亚细胞定位，从基因组序列数据中识别新的潜在可访问(表面暴露)的药物靶点和候选疫苗。在古原核生物中识别特定位置的蛋白质也越来越受到人们的关注，因为这种蛋白质具有潜在的重大工业应用(即耐热酶)和环境效用(检测具有重要生态意义的古菌)。此前，我领导了PSORTb的开发，它是目前世界上对细菌蛋白质亚细胞定位最准确的预测指标。这个程序，再加上我的团队开发的已知和计算预测的亚细胞定位蛋白质数据库，现在正被用来加速对具有特殊医疗、农业或环境兴趣的蛋白质的全基因组识别。然而，目前还没有在这种精度水平上对古生菌进行分析的计算预报器，因此我建议为此目的开发PSORTa。此外，虽然细菌PSORTb的精确度/特异性很高，但我假设，通过扩大与该程序相关的已知定位的蛋白质数据库，以及通过增加本提案中概述的新的计算分析，可以提高该方法的召回率/敏感度。我还建议，利用实验室分析参考细胞表面蛋白结构产生的数据，改进暴露在特定细胞表面蛋白上的序列的鉴定是可能的。由此产生的针对古生菌(PSORTa)和细菌(PSORTb)的亚细胞定位预测因子的扩展家族可用于对所有原核生物的蛋白质定位进行全球分析，从而对蛋白质定位和跨区域细胞网络的进化产生新的见解。这一改进的PSORT工具可以进一步加快对医疗、农业、环境和工业应用感兴趣的原核蛋白的鉴定。

There is a growing interest in both agriculture and medicine in computationally predicting the subcellular localization of bacterial proteins, to identify new potentially accessible (surface-exposed) drug targets and vaccine candidates from genomic sequence data. There is also a growing interest in the identification of proteins of particular localizations in archaeal prokaryotes, as such proteins have potential significant industrial applications (i.e. thermotolerant enzymes) and environmental utility (detection of ecologically important archaea). Previously, I led the development of PSORTb, which is currently the world's most precise predictor of bacterial protein subcellular localization. This program, coupled with a database my group developed of proteins of known and computationally-predicted subcellular localization, is now being used to accelerate genome-wide identification of proteins of particular medical, agricultural or environmental interest. However, there is currently no such computational predictor at this level of precision for the analysis of archaea and so I propose developing PSORTa for this purpose. In addition, while bacterial PSORTb precision/specificity is high, I hypothesize that the recall/sensitivity of this method could be improved though an expansion of the database of proteins of known localization associated with this program, and through the addition of a new computational analyses outlined in this proposal. I also propose that improved identification of sequences exposed on particular cell surface proteins is possible, using data generated from laboratory analysis of the structure of a reference cell-surface protein. The resulting expanded family of subcellular localization predictors, for both archaea (PSORTa) and bacteria (PSORTb), could be used to perform global analyses of protein localization across all prokaryotes, leading to new insights regarding the evolution of protein localization and cell networks across localizations. This improved PSORT tool could further accelerate the identification of prokaryotic proteins that are of interest for their medical, agricultural, environmental and industrial applications.