【摘要】 关键蛋白质是生物体存活和繁育后代所必需的蛋白质.现有方法主要通过在静态网络中引入时间等信息构建动态网络以提高关键蛋白质识别率.在从静态网络向动态网络构建的过程中,上述方法存在部分关键蛋白质缺失、相互作用信息冗余等缺点,影响关键蛋白质识别效率.提出一种新的动态蛋白质相互作用网络构建方法——全覆盖方法.该方法可保证静态网络的蛋白质相互作用信息全部映射到动态网络,并尽可能减少动态网络中蛋白质相互作用的重复率,即保证动态网络对静态网络的全覆盖并降低数据冗余度.实验表明,与静态网络和3Sigma动态网络相比,全覆盖动态网络在关键蛋白质识别方面具有优势.更多还原

【Abstract】 Essential proteins are important for the survival and reproduction of living things. The existing common method is to increase the recognition rate of key proteins by introducing time information in a static network and constructing a dynamic network. However, in the process of mapping from static network to dynamic network, there are some missing information of key protein and more redundant interaction information, which will affect the recognition efficiency of essential proteins. Aiming at the integrity of protein interaction data, a new dynamic protein network construction method full-overlapping method is proposed. Under the premise of ensuring the integrity of protein network data, this method can reduce the repeated protein interactions as much as possible and reduce the redundancy of data. The research shows that the full-overlapping dynamic protein network also has certain advantages in key protein recognition compared to fixed threshold methods and 3 Sigma methods.更多还原