Markov entropy backbone electrostatic descriptors for predicting proteins biological activity.

摘要: The spherical truncation of electrostatic interactions between aminoacids makes it possible to break down long-range spatial interactions, resulting in short-range interactions. As a result, Markov Chain model may be used calculate the probabilities with which effect given interaction reaches at different distances within backbone. entropies this type then codify information about distribution charges protein study exploring structure-activity relationship. In paper, linear discriminant analysis is reported, correctly classified 92.3% 26 under investigation training and leave-one-out cross validation, purely for illustrative purposes. Classification was carried out three activities: lysozymes, dihydrofolate reductases, alcohol dehydrogenases. equations were contracted into two canonical roots. These simple roots have high regression coefficients (R c1 = 0.903 R c2 0.70). Root1 explains biological activity dehydrogenases while Root2 discriminates lysozymes reductases. It profile core, middle, surface on activity. contrast, considering classic physicochemical parameters such as: polarizability, refractivity, partition coefficient classify only 80.8% proteins. © 2004 Elsevier Ltd. All rights reserved.