中文摘要：

把最大信息原理应用到核酸序列的保守位点分析中。利用最大信息原理，推导出了核酸和蛋白质特异性结合时的结合能表达式，并且估计了和蛋白质发生相互作用的核酸序列上的位点范围。为了检验此理论是否较为成功地反映了核酸和蛋白质结合时的实际情况，把它应用到基因内含子剪切位点的识别中，识别结果达到了较高的敏感性和特异性，这说明利用最大信息原理推导结合能表达式及估计核酸序列上参与反应的位点范围的理论是较为成功的。此研究结果一方面有助于核酸和蛋白质相互作用的理解，另一方面，也有助于和蛋白质发生相互作用的各种核酸序列的计算机识别研究。

英文摘要：

Maximum information principle was used to analyze the conservative positions in nucleic acid sequence. According to the maximum information principle, a binding energy expression was deduced, which represents the energy released when proteins recognize and bind specially with the nucleic acid sequence. On the other hand, a method was established on the basis of maximum information principle, which can be used to estimate site range reacting with proteins in nucleic acid sequence. To validate whether the theory is successful in explaining how nucleic acid and proteins bind specially, the authors applied it to recognize splicing sites in protein-coding genes. The result gives a high sensitivity and specificity, and it shows the theory can explain the recognizing and binding process of proteins with a nucleic acid sequence in splice sites successfully. The work also helps to the general problem of computational recognition of interacting sites between nucleic acids and proteins.