论文部分内容阅读
从20种天然氨基酸的41个分子轮廓指数(randic molecular profiles,R)、44个分子特征值指数(eigenvalue based indices,E)和47个分子运转路径数目(walk and path counts,W)分别进行主成分分析,得出1种新的氨基酸描述符(scores vector of R,E,W-SVREW)。将其应用于血管紧张素转化酶(ACE)抑制二肽、三肽、四肽、九肽结构表征,应用多元线性回归建立定量构效关系模型,同时采用内部与外部双重验证的方法验证模型的稳定性。所建ACE抑制二肽、三肽、四肽、九肽的模型复相关系数(R_(cum)~2)、留一法(LOO)交互校验复相关系数(R_(cv)~2)和外部样本校验相关系数(Q_(ext)~2)分别为:0.907、0.791、0.633;0.831、0.603、0.723;0.834、0.668、0.718;0.964、0.853、0.948。经研究表明:SVREW描述符应用于ACE抑制肽结构表征所建模型稳定性与预测能力均较好。
From 41 molecular profiles of 20 natural amino acids (R), 44 molecular eigenvalue based indices (E) and 47 molecular weights of walk and path counts (W) Component analysis resulted in a new kind of scores vector of R (E, W-SVREW). It was applied to the inhibition of dipeptide, tripeptide, tetrapeptide and nonapeptide structures by angiotensin converting enzyme (ACE), and the quantitative structure-activity relationship model was established by multiple linear regression. The internal and external double verification methods were used to verify the model stability. The correlation coefficient (R_ (cum) ~ 2), the LOV cross validation (R_ (cv) ~ 2) of ACE inhibit dipeptide, tripeptide, tetrapeptide and nonapeptide were The correlation coefficient of external sample (Q ext ~ 2) is 0.907,0.791,0.633; 0.831,0.603,0.723; 0.834,0.668,0.718; 0.964,0.853,0.948, respectively. The study shows that: SVREW descriptors used in ACE inhibitory peptide structural characterization model established model stability and prediction are good.