用GEP实现复杂函数的自动建模

GEP是一种新颖的遗传算法,在函数建模的应用中取得良好的结果.给出计算有效基因长度的伪代码,结合GRCM方法阅读基因,快速计算出染色体的适应值.在算法中增加了参数估计模块,用GEP得到较好模型后,用参数估计模块进行参数优化,试验显示这种混合的GEP方法比传统的最小二乘法、神经网络以及遗传程序设计等方法具有更好的性能.     

Mechanisms of sperm-egg interactions emerging from gene-manipulated animals

Untangling the molecular nature of sperm-egg interactions is fundamental if we are to understand fertilization. These phenomena have been studied for many years using biochemical approaches such as antibodies and ligands that interact with sperm or with eggs and their vestments. However, when homologous genetic recombination techniques were applied, most of the phenotypic factors of the gene-manipulated animals believed “essential” for fertilization were found to be dispensable. Of course, all biological systems contain redundancies and compensatory mechanisms, but as a whole the old model of fertilization clearly requires significant modification. In this review, we use the results of gene manipulation experiments in animals to propose the basis for a new vision. Received 26 January 2007; received after revision 7 March 2007; accepted 17 April 2007     

The human genome and understanding of common disease: present and future technologies

The study of candidate genes over the past three decades has yielded notable successes in common-disease genetics. During this time, however, interpretation of genetic association studies has been hampered by the use of clinical cohorts of inadequate power and insufficient information on genetic variation in candidate genes. The unavailability of highthroughput and low-cost genotyping technologies has also limited the scope of complex-disease genetic studies. More recently, however, the sequencing and characterization of variation within the human genome has revolutionized genetic studies and enabled full genome-wide scans for genes associated with disease. The identification of disease-associated (causative) genes has illuminated disease mechanisms. The translation of this knowledge into direct clinical benefit in diagnosis, prognosis and therapy for an individual’s disease still remains a challenge. Received 11 September 2006; received after revision 17 December 2006; accepted 18 January 2007     

中国四个小型猪品种生长激素基因克隆测序及分子进化研究

研究对中国四个小型猪五指山猪、贵州香猪、滇南小耳猪和藏猪的生长激素基因(pGH,porcine growth hormone)进行了克隆测序及构建分子进化树,考察该激素对小型猪体型的影响。通过筛选合适的引物,采用PCR技术,扩增了四个小型猪品种的pGH基因全序列,并对其进行了克隆测序分析。4个小型猪品种pGH基因全长为2006bp,包括5个外显子和4个内含子,CDS全长为648bp。将4个品种小型猪和长白猪、雅南猪、内江猪进行了核苷酸序列比对,共有63处发生了变异,变异率为2.9%,其中外显子有12处变异,全部为转换;内含子有51处发生了变异,包括转换、颠换和缺失。聚类结果基本符合其地方猪种的地理位置分布原则。     

血管生成抑制因子arresten克隆及功能的初步研究

血管生成抑制因子arresten是血管生成及肿瘤生长抑制物.研究发现,将arresten基因重组于质粒pGEX4T1转化入大肠杆菌BL21并用IPTG诱导蛋白表达,可获得分子量为45kDa的目的蛋白条带,arresten蛋白多数以包涵体形式存在,也有可溶蛋白存在于菌裂解液的上清,可溶蛋白可由Sepharose 4B凝胶柱亲和层析柱纯化.在鸡胚绒毛尿囊膜(chick chorioallantoic membrane,CAM)实验中,arresten基因可溶蛋白被证实抑制新生血管生成的效果较好.     

基因预测中外显子分值的转化

多数基因预测算法都会给出预测外显子的分值,但这些分值大多没有意义,很难给使用者带来有用的信息.作者以基因预测程序Fgenes为例,分别用4种方法将外显子的分值转换成概率分值,通过计算机模拟,4种方法都取得了良好的效果,其中以局部多项式估计效果最好.     

基于基因调控的改进遗传算法

针对标准遗传算法存在个体多样性差,易陷入局部极小等缺陷,提出了基于基因调控的改进遗传算法。该方法具有结构简单、搜索速度快、操作方便等特点。典型函数测试实验及对比分析结果证明了该方法的有效性,为遗传算法的改进及智能优化方法的研究提供了一条新的途径。     

压电生物传感器的研究进展

压电生物传感器是一种将高灵敏的压电传感器与特异的生物反应结合在一起的新型生物分析方法,这一方法不需要任何标记,且仪器构造简单、操作方便,引起人们的兴趣,并已成为当前生物传感器领域中的一个研究热点。本文就压电生物传感器在气体、微生物、病毒、蛋白质及基因检测等方面的研究应用作一评述。