Two-level QSAR network (2L-QSAR) for peptide inhibitor design based on amino acid properties and sequence positions
Two-level QSAR network (2L-QSAR) for peptide inhibitor design based on amino acid properties and sequence positions
复制标题
基于氨基酸特性和序列位置的肽抑制剂设计的两级 QSAR 网络 (2L-QSAR)。
DOI：
10.1080/1062936x.2014.959049
复制
发表时间：
2014-10-03
影响因子：
3
通讯作者：
Huang, R. B.
中科院分区：
文献类型：
--
作者：
Du, Q. S.;Ma, Y.;Huang, R. B.
在多肽抑制剂的设计中，多肽序列的巨大可能的多样性是高度关注的。结合快速积累的多肽实验数据和数据库，提出了多肽抑制剂设计的统计方法。在两级肽预测网络(2L-QSAR)中，一级是氨基酸的物理化学性质，另一级是肽序列的位置。氨基酸的活性贡献是理化性质和序列位置的函数。在预测方程中，将两个权重系数集{a(K)}和{b(L)}分别赋予物化性质和序列位置。根据已知多肽抑制剂的实验数据，利用迭代双最小二乘法(IDLS)对这两个系数集进行优化后，用这些系数来评价新设计的多肽抑制剂的生物活性。该两级预测网络可应用于针对不同靶蛋白或蛋白质不同位置的多肽抑制剂设计。两级统计算法的一个显著优点是不需要宿主蛋白质的结构信息。它还可以提供对氨基酸性质和序列位置的作用的有用的见解。
In the design of peptide inhibitors the huge possible variety of the peptide sequences is of high concern. In collaboration with the fast accumulation of the peptide experimental data and database, a statistical method is suggested for peptide inhibitor design. In the two-level peptide prediction network (2L-QSAR) one level is the physicochemical properties of amino acids and the other level is the peptide sequence position. The activity contributions of amino acids are the functions of physicochemical properties and the sequence positions. In the prediction equation two weight coefficient sets {a(k)} and {b(l)} are assigned to the physicochemical properties and to the sequence positions, respectively. After the two coefficient sets are optimized based on the experimental data of known peptide inhibitors using the iterative double least square (IDLS) procedure, the coefficients are used to evaluate the bioactivities of new designed peptide inhibitors. The two-level prediction network can be applied to the peptide inhibitor design that may aim for different target proteins, or different positions of a protein. A notable advantage of the two-level statistical algorithm is that there is no need for host protein structural information. It may also provide useful insight into the amino acid properties and the roles of sequence positions.