Molecularly imprinted polymers(MIPs) are artificial, template-made receptors with the ability to recognize and to specially bind the target molecule. The advantage of stability of the polymer, ease of the preparation ...Molecularly imprinted polymers(MIPs) are artificial, template-made receptors with the ability to recognize and to specially bind the target molecule. The advantage of stability of the polymer, ease of the preparation and low cost of these MIPs have led to their assessment as substitutes for antibodies or enzymes in chemical sensors, catalysis and separations. Although creating a MIPs against small molecules is straightforward now, imprinting of large structures, such as proteins and other biomacromolecules, is still a challenge. The major problem associated with the imprinting of such large structures lies on the restricted mobility of them within highly cross-linked polymer networks and the poor efficiency in rebinding. In this paper, we present a technique for the preparation of polymer nanowires with the protein molecule imprinted and binding sites on surface. These surface imprinting nanowires exhibit highly selective recognition for a variety of template proteins, including albumin, hemoglobin and cytochrome c. Since the protein imprinted sites are located on, or close to, the surface, these imprinted nanowires have a good site accessibility towards the target protein molecules. Furthermore, the large surface area of the nanowires results in larger protein molecules binding capacity of the imprinted nanowires compared to previously report surface imprinting MIPs.展开更多
蛋白质关系抽取研究对于生命科学各领域的研究具有广泛的应用价值。但是,基于机器学习的蛋白质关系抽取方法普遍停留在二元关系抽取,失去了丰富的关系类型信息,而基于规则的开放式信息抽取方法可以抽取完整的蛋白质关系("蛋白质1,...蛋白质关系抽取研究对于生命科学各领域的研究具有广泛的应用价值。但是,基于机器学习的蛋白质关系抽取方法普遍停留在二元关系抽取,失去了丰富的关系类型信息,而基于规则的开放式信息抽取方法可以抽取完整的蛋白质关系("蛋白质1,关系词,蛋白质2"),但是召回率较低。针对以上问题,该文提出了一种混合机器学习和规则方法的蛋白质关系抽取框架。该框架先利用机器学习方法完成命名实体识别和二元关系抽取,然后利用基于句法模板和词典匹配的方法抽取表示当前两个蛋白质间关系类型的关系词。该方法在AImed语料上取得了40.18%的F值,远高于基于规则的Stanford Open IE方法。展开更多
蛋白质复合物是许多生物过程得以实现的基石。蛋白质相互作用数据中的假阳性和假阴性对各种识别蛋白质复合物的计算方法有不良影响。为了解决这一问题,1种新的蛋白质复合物识别算法(ICMDS,Identifying Complexes based on Multiple Data...蛋白质复合物是许多生物过程得以实现的基石。蛋白质相互作用数据中的假阳性和假阴性对各种识别蛋白质复合物的计算方法有不良影响。为了解决这一问题,1种新的蛋白质复合物识别算法(ICMDS,Identifying Complexes based on Multiple Data Sources)被提出。该方法整合基因表达谱、关键蛋白质信息和蛋白质相互作用3种生物数据进行蛋白质复合物的挖掘。首先,ICMDS重新定义了2个相互作用的蛋白质之间的功能相似性(FS,Functional Similarity)。然后,ICMDS选择已知的关键蛋白质作为种子构建蛋白质复合物。为了消除冗余的复合物,ICMDS算法也设计了冗余过滤子程序。另外,ICMDS也使用非关键蛋白质作为种子并将之扩展为蛋白质复合物。实验结果表明ICMDS识别蛋白质复合物的能力明显优于其他计算方法。展开更多
文摘Molecularly imprinted polymers(MIPs) are artificial, template-made receptors with the ability to recognize and to specially bind the target molecule. The advantage of stability of the polymer, ease of the preparation and low cost of these MIPs have led to their assessment as substitutes for antibodies or enzymes in chemical sensors, catalysis and separations. Although creating a MIPs against small molecules is straightforward now, imprinting of large structures, such as proteins and other biomacromolecules, is still a challenge. The major problem associated with the imprinting of such large structures lies on the restricted mobility of them within highly cross-linked polymer networks and the poor efficiency in rebinding. In this paper, we present a technique for the preparation of polymer nanowires with the protein molecule imprinted and binding sites on surface. These surface imprinting nanowires exhibit highly selective recognition for a variety of template proteins, including albumin, hemoglobin and cytochrome c. Since the protein imprinted sites are located on, or close to, the surface, these imprinted nanowires have a good site accessibility towards the target protein molecules. Furthermore, the large surface area of the nanowires results in larger protein molecules binding capacity of the imprinted nanowires compared to previously report surface imprinting MIPs.
文摘蛋白质关系抽取研究对于生命科学各领域的研究具有广泛的应用价值。但是,基于机器学习的蛋白质关系抽取方法普遍停留在二元关系抽取,失去了丰富的关系类型信息,而基于规则的开放式信息抽取方法可以抽取完整的蛋白质关系("蛋白质1,关系词,蛋白质2"),但是召回率较低。针对以上问题,该文提出了一种混合机器学习和规则方法的蛋白质关系抽取框架。该框架先利用机器学习方法完成命名实体识别和二元关系抽取,然后利用基于句法模板和词典匹配的方法抽取表示当前两个蛋白质间关系类型的关系词。该方法在AImed语料上取得了40.18%的F值,远高于基于规则的Stanford Open IE方法。
文摘蛋白质复合物是许多生物过程得以实现的基石。蛋白质相互作用数据中的假阳性和假阴性对各种识别蛋白质复合物的计算方法有不良影响。为了解决这一问题,1种新的蛋白质复合物识别算法(ICMDS,Identifying Complexes based on Multiple Data Sources)被提出。该方法整合基因表达谱、关键蛋白质信息和蛋白质相互作用3种生物数据进行蛋白质复合物的挖掘。首先,ICMDS重新定义了2个相互作用的蛋白质之间的功能相似性(FS,Functional Similarity)。然后,ICMDS选择已知的关键蛋白质作为种子构建蛋白质复合物。为了消除冗余的复合物,ICMDS算法也设计了冗余过滤子程序。另外,ICMDS也使用非关键蛋白质作为种子并将之扩展为蛋白质复合物。实验结果表明ICMDS识别蛋白质复合物的能力明显优于其他计算方法。
