ACE抑制肽定量构效关系研究

VIP免费

3.0 侯斌 2024-11-19 4 4 1.84MB 82 页 15积分

侵权投诉

摘要

近年来高血压发病率呈上升趋势，造成了沉重的家庭和社会负担。食源性血

管紧张素转化酶（angiotensin I-converting enzyme,ACE）抑制肽可通过抑制生物体

内ACE 活性起到良好的降压效果，但具体抑制机制尚不明确。定量构效关系

（quantitative structure activity relationship,QSAR）研究能提供分子结构与活性的定

量模型，对 ACE 抑制肽进行 QSAR 研究可为深入探讨 ACE 抑制肽的作用机制，

以及设计和开发高效降血压药物提供指导作用。

利用 20 种天然氨基酸的 457 种物化参数构建了新的氨基酸结构描述符

SVHEHS，并运用两个经典肽库对其效果进行评价。58 个ACE 抑制二肽的多元线

性回归（multivariate linear regression,MLR）模型相关系数（R2）为 0.936，均方根

误差（RMSE）为 0.259，交叉验证相关系数（Q2LOO）为 0.854，外部验证相关系

数（Q2ext）为 0.737；

48 个苦味二肽的 MLR 模型 R2为0.949，

RMSE 为0.136，

Q2LOO

为0.886，Q2ext 为0.543。与其他氨基酸描述符相比，SVHEHS 描述符建立的模型

在拟合能力、稳健性及预测能力上都更优，说明该描述符可用于对生物活性肽进

行定量构效关系研究。

利用 SVHEHS 描述符和 MLR 分别对自组建 ACE 抑制二、三、四肽进行 QSAR

研究。结果表明，二肽模型 R2=0.851，RMSE=0.327，Q2LOO=0.781，Q2ext=0.792，

且C端氨基酸残基疏水性质及电荷性质和 N端氨基酸残基立体特征对 ACE 抑制二

肽的活性影响较大，特别是 C端氨基酸残基强的疏水性和弱的电荷性质对其活性

有积极作用；三肽模型 R2=0.805，RMSE=0.339，Q2LOO=0.717，Q2ext=0.817，N端

第一位氨基酸残基的疏水性、电性特征、立体特征及 C端氨基酸残基的疏水性、

电性特征对肽的活性影响较大，尤以第一位氨基酸残基疏水性影响较为显著；四

肽模型 R2=0.792，RMSE=0.393，Q2LOO=0.553，Q2ext=0.630，四肽从 N端开始第三

位氨基酸残基的氢键贡献和电性特征对 ACE 抑制四肽的活性影响较大，尤以第三

位氨基酸残基氢键贡献与活性呈显著负相关；对三种肽合集的建模结果为：

R2=0.744，RMSE=0.508，Q2LOO=0.532，Q2ext=0.567。

利用 SVHEHS 描述符和偏最小二乘（partial least square regression,PLS）对自

组建 ACE 抑制二、三、四肽进行 QSAR 研究。结果表明，二肽模型 R2=0.607，

RMSE=0.587，Q2LOO=0.507，Q2ext=0.783，C端氨基酸残基的疏水性、电性特征、

立体特征和 N端氨基酸残基的立体特征与肽活性相关性较大；三肽模型 R2=0.852，

RMSE=0.232，Q2LOO=0.813，Q2ext=0.839，N端氨基酸残基的疏水性、立体特征和

C端氨基酸残基的立体、电性特征对活性有显著影响；四肽模型 R2=1，RMSE=0，

Q2LOO=1，Q2ext=0.935，从 N端开始第三位氨基酸残基的电性，第二位氨基酸残基

的立体特征，C端氨基酸残基的立体特征和 N端氨基酸的氢键贡献对肽活性影响

较大；对肽合集的建模结果为 R2=0.862，RMSE=0.356，Q2LOO=0.829，Q2ext=0.632。

利用 SVHEHS 描述符和人工神经网络（artificial neural networks,ANN）对自组

建ACE 抑制二、三、四肽进行 QSAR 研究。结果表明，二肽模型 R2=0.946，

RMSE=0.249，Q2LOO=0.951，Q2ext=0.852，N、C端氨基酸残基立体特征对肽活性

影响显著；三肽模型 R2=0.973，RMSE=0.135，Q2LOO=0.945，Q2ext=0.813，影响三

肽活性的主要因素为 N端氨基酸残基疏水性、立体特征和 C端氨基酸残基立体特

征等；四肽模型 R2=0.915，RMSE=0.250，Q2LOO=0.879，Q2ext=0.814，影响四肽活

性的主要因素为 C端氨基酸残基立体特征、第二位氨基酸残基电性特征以及第三

位氨基酸残基电性特征等；对肽合集的建模结果为：R2=0.958，RMSE=0.224，

Q2LOO=0.948，Q2ext=0.634。

最后，本文对 QSAR 建模的 3种算法优劣进行了比较，结果表明 MLR 能得到

明确的函数表达式，有利于解释和分析模型，但对变量正交性要求较严；PLS 对

小样本容量、多自变量个数且变量之间严重多重相关的肽库有很大优势，但拟合

能力稍差；ANN 能拟合非线性关系，快速逼近最佳样本数据规律，但无法获得明

确的函数关系式，且网络隐含层数和神经元节点个数没有确定规则，网络的学习

和记忆欠稳定。尽管 3种算法对自建肽库拟合能力有差异，但模型均能给出影响

肽活性的共性结构特征。影响二肽活性的结构主要是 C端氨基酸残基的疏水性和

N端氨基酸残基的立体特征；影响三肽活性的结构主要是 N端氨基酸残基的疏水

性和立体特征；影响四肽活性的结构主要是第三位氨基酸的电性特征。将不同长

度肽融合在一个模型，能减少建模时间，为预测更多不同长度的肽活性提供便利。

关键词：血管紧张素转化酶抑制肽定量构效关系氨基酸描述符主成

分分析多元线性回归偏最小二乘人工神经网络

ABSTRACT

In recent years the incidence of hypertension has been rising, and it has resulted in

heavy burden on society and family. Angiotensin I-converting enzyme(ACE) inhibitory

peptide from food protein can inhibit ACE activity in vivo significantly,but the specific

mechanism is unclear.Quantitative structure activity relationship(QSAR) research is

useful to establish quantitative model between the molecular structure and activity.

QSAR research on ACE inhibitory peptide is helpful for looking into action mechanism

of ACE inhibitory peptide,and then guiding the design and development of effective

hypotensor.

SVHEHS descriptor was derived from 457 physicochemical properties indexes of

20 natural amino acids,and its performance was tested by 58 angiotensin I-converting

enzyme inhibitory dipeptides and 48 bitter taste dipeptides.The correlative coefficient

R2,root mean square error RMSE,the cross-validation correlative coefficient Q2LOO and

the external validation correlation coefficient Q2ext of two models were 0.936, 0.259,

0.854, 0.737; 0.949, 0.136, 0.886 and 0.543,respectively.Compared the above statistics

parameters of models with other models constructed by various descriptors,the fitting

ability,robustness and forecasting ability of SVHEHS were better.These showed that the

descriptor can be used for QSAR research.

SVHEHS descriptor and multiple linear regression algorithm were used to study

the QSAR of angiotensin I-converting enzyme inhibitory dipeptides,tripeptides and

tetrapeptides.The parameters R2, RMSE, Q2LOO, Q2ext of dipeptides model were 0.851,

0.327, 0.781 and 0.792,respectively.Hydrophobicity and charge of C-terminal amino

acid,and steric properties of N-terminal amino acid were closely related with its activity.

Hydrophobicity of C-terminal amino acid was most positively associated with the

activity,and charge of C-terminal amino acid most negatively with the activity.The

parameters R2, RMSE, Q2LOO, Q2ext of tripeptides model were 0.805, 0.339, 0.717 and

0.817,respectively.Hydrophobic,electronic and steric properties of N-terminal amino

acid,hydrophobic and electronic properties of C-terminal amino acid were closely

related with its activity,especially the hydrophobicity of N-terminal amino acid was

significant factor with positive effect.The parameters R2, RMSE, Q2LOO, Q2ext of

tetrapeptides model were 0.792, 0.393, 0.553 and 0.630,respectively.Hydrogen bonds

contributions and electrical characteristics of the third amino acid were closely related

with its activity,especially hydrogen bonds contributions was significant factor with

negative effect.The parameters R2, RMSE, Q2LOO, Q2ext of all of the peptides model were

0.744, 0.508, 0.532 and 0.567,respectively.

SVHEHS descriptor and partial least square algorithm were used to study the

QSAR of angiotensin I-converting enzyme inhibitory dipeptides, tripeptides and

tetrapeptides.The parameters R2, RMSE, Q2LOO, Q2ext of dipeptides model were 0.607,

0.587, 0.507 and 0.783,respectively.Hydrophobic,electronic and steric properties of

C-terminal amino acid,steric features of N-terminal amino acid were closely related

with its activity.The parameters R2, RMSE, Q2LOO, Q2ext of tripeptides model were 0.852,

0.232, 0.813 and 0.839,respectively.Hydrophobic and steric properties of N-terminal

amino acid, electronic and steric properties of C-terminal amino acid were closely

related with its activity.The parameters R2, RMSE, Q2LOO, Q2ext of tetrapeptides model

were 1, 0, 1 and 0.935,respectively.Electronic properties of the third amino acid,steric

properties of the second amino acid,steric properties of the C-terminal amino

acid,hydrogen bonds contributions of N-terminal amino acid were closely related with

its activity.The parameters R2, RMSE, Q2LOO, Q2ext of all of the peptides model were

0.862, 0.356, 0.829 and 0.632,respectively.

SVHEHS descriptor and artificial neural networks algorithm were used to study

the QSAR of angiotensin I-converting enzyme inhibitory dipeptides,tripeptides and

tetrapeptides.The parameters R2, RMSE, Q2LOO, Q2ext of dipeptides model were 0.946,

0.249, 0.951 and 0.852,respectively.Steric features of N-terminal amino acid and

C-terminal amino acid were closely related with its activity.The parameters R2, RMSE,

Q2LOO, Q2ext of tripeptides model were 0.973, 0.135, 0.945 and 0.813, respectively.

Hydrophobic and steric properties of N-terminal amino acid, steric properties of

C-terminal amino acid were closely related with its activity. The parameters R2, RMSE,

Q2LOO, Q2ext of tetrapeptides model were 0.915, 0.250, 0.879 and 0.814, respectively.

Steric features of C-terminal amino acid,electronic properties of the second amino acid,

electronic properties of the third amino acid were closely related with its activity. The

parameters R2, RMSE, Q2LOO, Q2ext of all of the peptides model were 0.958, 0.224,

0.948 and 0.634, respectively.

Finally,advantages and disadvantages of the three algorithms were compared.

Multiple linear regression can obtained clear function expressions, and be helpful for

model interpretation and analysis. However, its requirements of variable orthogonality

was strict. Partial least squares had great advantage to small samples with many

independent variables and serious related data, but it could not fit all of the database

well. Artificial neural network can fit nonlinear relation, and approach the best sample

data law quickly, but it had no explicit function equation. Moreover, there were

uncertain rules for the network layers and nodes, learning and memory of the network

were less stable than the former two algorithms. Although fitting abilities of three

algorithms to peptide library varied from each other, models drew the common structure

features which affect the peptides activities. Hydrophobicity and charge of C-terminal

amino acid,steric properties of N-terminal amino acid were more associated with

dipeptides activity. Hydrophobic and steric properties of N-terminal amino acid were

associated with tripeptides activity much. Charge of the third amino acid was associated

with tetrapeptides activity much.Models, which established by peptides with different

length, can make a prediction for the activities of different peptides in one model. Hence,

it can consume less time for modeling. This method can facilitate activities prediction of

different peptides when the impact factors on activities were clear.

Key Words ：angiotensin I-converting enzyme inhibitory peptides,

quantitative structure activity relationship,amino acid descriptor,

principal component analysis, multiple linear regression,partial least

square regression,artificial neural networks

中文摘要

ABSTRACT

第一章绪论........................................................................................................1

§1.1 血管紧张素转化酶（ACE）与高血压.....................................................1

§1.1.1 ACE 概述 ............................................................................................1

§1.1.2 ACE 在血压调节中的作用..................................................................1

§1.2 ACE 抑制剂.................................................................................................2

§1.3 定量构效关系（QSAR）..........................................................................3

§1.3.1 二维定量构效关系(2D-QSAR).......................................................... 4

§1.3.2 三维定量构效关系(3D-QSAR).......................................................... 5

§1.3.3 QSAR 研究方法...................................................................................5

§1.4 ACE 抑制肽 QSAR 研究进展.................................................................... 8

§1.5 研究意义和内容.......................................................................................10

§1.5.1 研究意义............................................................................................10

§1.5.2 研究内容............................................................................................11

第二章氨基酸描述符 SVHEHS 的建立与效果评价...................................... 12

§2.1 原理和方法...............................................................................................12

§2.1.1 SVHEHS 的建立................................................................................ 12

§2.1.2 SVHEHS 的效果评价........................................................................ 14

§2.2 结果与分析...............................................................................................14

§2.2.1 经典 ACE 抑制二肽库用于 SVHEHS 的评价................................ 14

§2.2.2 经典苦味二肽库用于 SVHEHS 的评价.......................................... 17

§2.3 结论...........................................................................................................19

第三章 MLR 用于 ACE 抑制肽 QSAR 研究................................................... 21

§3.1 原理和方法...............................................................................................21

§3.1.1 ACE 抑制肽肽库的建立....................................................................21

§3.1.2 SVHEHS 描述符用于肽结构表征.................................................... 21

§3.1.3 MLR 建模与模型检验.......................................................................21

§3.2 结果与分析...............................................................................................22

§3.2.1 ACE 抑制二肽的 QSAR 研究........................................................... 22

§3.2.2 ACE 抑制三肽的 QSAR 研究........................................................... 24

§3.2.3 ACE 抑制四肽的 QSAR 研究........................................................... 28

文档加载中……请稍候！
如果长时间未打开，您也可以点击刷新试试。

下载文档到电脑，查找使用更方便

15 积分 4人已下载

立即下载 VIP免费下载

摘要：

摘要近年来高血压发病率呈上升趋势，造成了沉重的家庭和社会负担。食源性血管紧张素转化酶（angiotensinI-convertingenzyme,ACE）抑制肽可通过抑制生物体内ACE活性起到良好的降压效果，但具体抑制机制尚不明确。定量构效关系（quantitativestructureactivityrelationship,QSAR）研究能提供分子结构与活性的定量模型，对ACE抑制肽进行QSAR研究可为深入探讨ACE抑制肽的作用机制，以及设计和开发高效降血压药物提供指导作用。利用20种天然氨基酸的457种物化参数构建了新的氨基酸结构描述符SVHEHS，并运用两个经典肽库对其效果进行评价。5...

展开>> 收起<<

ACE抑制肽定量构效关系研究.pdf

共82页,预览9页

还剩页未读，继续阅读

ACE抑制肽定量构效关系研究

相关推荐

开通VIP享超值会员特权

作者详情

相关内容

推荐作者

热门标签

举报选择: