Georgiou D N, Karakasidis T E, Nieto J J, Torres A
Department of Mathematics, University of Patras, 265 00 Patras, Greece.
J Theor Biol. 2009 Mar 7;257(1):17-26. doi: 10.1016/j.jtbi.2008.11.003. Epub 2008 Nov 12.
In this paper we present a study of classification of the 20 amino acids via a fuzzy clustering technique. In order to calculate distances among the various elements we employ two different distance functions: the Minkowski distance function and the NTV metric. In the clustering procedure we take into account several physical properties of the amino acids. We examine the effect of the number and nature of properties taken into account to the clustering procedure as a function of the degree of similarity and the distance function used. It turns out that one should use the properties that determine in the more important way the behavior of the amino acids and that the use of the appropriate metric can help in defining the separation into groups.
在本文中，我们展示了一项通过模糊聚类技术对20种氨基酸进行分类的研究。为了计算各种元素之间的距离，我们采用了两种不同的距离函数：闵可夫斯基距离函数和NTV度量。在聚类过程中，我们考虑了氨基酸的几种物理性质。我们研究了所考虑的性质的数量和性质对聚类过程的影响，将其作为相似度和所用距离函数的函数。结果表明，应该使用以更重要方式决定氨基酸行为的性质，并且使用适当的度量有助于定义分组。
