论文部分内容阅读
生物系统的精确盲识别是生物科学研究的核心问题之一。采用生物相似样品统计判别规则P≥P珔+xS,确定每个样品个性化特征序列;建立N个相似生物体系准确盲识别的通用判别标准Mijc>ml,l≠j,并确定每个样品的核心特征序列。每类样品核心特征序列中的样品构成独立的集合,构建了N-生物系统双指标等级序列个性化模式识别法,从而对多种生物复杂系统进行精确盲分类识别。采用该法分析了3种组成相似的复杂生物体系左归丸、右归丸、济生肾气丸无水乙醇提取物的1H-NMR指纹图谱,样品的准确识别率达92%。利用该理论方法,既可以对复杂生物系统进行精确鉴别,又可利用关联序列精确研究生物系统之间的关联关系,为生物演化研究提供了一种可行的理论方法。
Accurate blind identification of biological systems is one of the core issues in biological sciences. The statistical discriminant rule of biosimilar sample is defined as P≥P 珔 + xS, and the personalized feature sequence of each sample is determined. A general discriminant criterion Mijc> ml, l ≠ j for accurate blind identification of N similar biological systems is established, Core feature sequence. The samples of each type of sample core feature sequence form an independent set, and a personalized pattern recognition method of bi-indicator rank sequence of N-biological system is constructed to identify the precise biological classification system of complex biological systems. Using this method, 1H-NMR fingerprints of three kinds of complex biological systems with similar composition, such as Zuogui pill, Yougui pill and Jisheng Shenqi pill, were analyzed. The accurate identification rate of the samples was 92%. The method can not only accurately identify complex biological systems, but also use correlation sequences to accurately study the relationships between biological systems, providing a feasible theoretical method for biological evolution research.