Strategy for the mapping of interactive genes using bulked segregant analysis method and Mapmaker/Exp software

A qualitative trait is usually controlled by a single gene, but it may be sometimes controlled by two or even more genes. This phenomenon is called gene interaction. Rapidly searching for linked mo- lecular markers via bulked segregant analysis (BSA) and then constructing regional linkage map with Mapmaker/Exp has become a common approach to mapping single major genes. However, methods and computer programs developed for mapping single major genes cannot be simply applied to interactive genes because the genetic patterns of gene interac- tions are quite different from that of single-gene in- heritance. Up to now, experimental methods for quickly screening molecular markers linked to inter- active genes and statistical methods and corre- sponding computer softwares for simultaneously analyzing the linkage relationships of multiple mo- lecular markers to an interactive gene have not been available. To solve this problem, in this paper, we propose a strategy for mapping interactive genes using BSA and Mapmaker/Exp. We demonstrate that all interactive genes can be mapped by the 'BSA Mapmaker/Exp' strategy using F2 generation (in a few cases, F3 generation is also needed). As BSA and Mapmaker/Exp have been broadly used in gene mapping studies and are well known by many re- searchers, the strategies proposed in this paper will be useful for practical researches.     

Genetic analysis and molecular mapping of a presenescing leaf gene psll in rice （Oryza sativa L.）

A rice psl1 （presenescing leaf） mutant was obtained from a japonica variety Zhonghua 11 via radiation of ^60Co-γ in M2 generation. Every leaf of the mutant began to wither after it reached the biggest length, while the leaves of the wild variety could keep green for 25--35 d. In this study, genetic analysis and gene mapping were carried out for the mutant identified. The SSR marker analysis showed that the mutant was controlled by a single recessive gene （psl1） located on chromosome 2. Fine mapping of the psl1 locus was conducted with 34 new STS markers developed around psl1 anchored region based on the sequence diversity between Nipponbare and 93-11. The psl1 was further mapped between two STS markers, STS2-19 and STS2-26, with genetic distances of 0.43 and 0.11 cM, respectively, while cosegregated with STS2-25. A BAC contig was found to span the psl1 locus, the region being delimited to 48 kb. This result was very useful for cloning of the psl1 gene.     

局部核方法及其应用

核方法是机器学习中一种强有力的学习算法.多个领域的实践表明，通过将领域知识嵌入到核函数中，一般会得到比较好的学习效果. 从微分流形的观点讨论了机器学习中全局信息与局部信息的关系，并提出了一种嵌入局部有意义信息的核方法.文本分类的实验结果表明，与其它几个分类算法相比，它具有较高的性能.     

Banach空间中增生型变分包含解的具误差的Ishikava序列逼近问题

在Banach空间中研究一类增生型变分包含解的存在性及其具误差的Ishikava迭代程序的收敛性问题,对文献中相应结果进行了改进与推广.     

基于C++的实时数据库的设计与实现

根据仿真系统支撑平台的实时性要求，利用动态链接库和内存映射文件建立系统运行的实时数据库，通过内存映射文件开辟共享内存区，编制一系列封装在动态链接库的接口函数来实现对共享内存区数据的操作．该方法既能实现多个进程数据的共享，又保证了实时的响应速度．     

关于弱J-空间的一些性质

目的　研究弱J 空间、半弱J 空间及其乘积空间和J 空间的锥空间的一些性质。方法　在前人研究J 空间的基础之上,用类比的思想以及构造新空间(乘积空间与商空间)的方法来研究与弱J 空间有关的一些性质。结果　得到了有关弱J 空间和半弱J 空间的一些等价命题及乘积性质,也得到了J 空间的锥空间的性质。结论　①设{X1,X2,K}是空间X的闭覆盖并且满足K紧,X1∩X2= ,则X是弱J 空间当且仅当X1和X2是弱J 空间,且X1或X2紧;②判断乘积空间X1×X2是弱J 空间的一些充分必要条件;③如果X是连通的J 空间,那么Δ(X)是半弱J 空间。     

有限集上的一类映射的计数

设A是有限集,|A|=n,F(n,m)是从A到A的映射中满足fm(x)=f(x)的映射f的个数,在此给出了计数函数F(n,m)的一个表达式.     

视觉核感知器

将Mercer核思想和视觉理论相结合，提出了视觉核感知器算法(VKP)。相对于单纯的核感知器，视觉核感知器能更有效地解决了学习的收敛性问题。同时也有效地提高了核感知器学习的精度。大量的实验结果表明了本算法的有效性和某些情况下的优越性。     

嵌入式Linux下网络设备驱动程序的开发

阐述了Linux网络系统的分层结构，分析了Linux内核中网络设备驱动程序的运行机理，介绍了实现Linux网络驱动程序开发的关键过程，结合设备驱动的调试方法通过应用实例给出了开发的全过程．最后，总结了嵌入式Linux下网络设备驱动程序开发的通用实现模式．     

板材冲裁和翻口工艺特征识别及图形转换的研究

该文针对冲孔、翻孔两类板材冲压成形工艺方法的主要特点，设计了两类成形件的特征识别规则．根据形状特征与冲压加工特征间存在着单映射的关系，可以充分利用模型中的几何信息和拓扑关系来识别零件的特征．因板材零件的厚度与整个零件尺寸比较是小的，可以假设零件厚度为常数．在消除厚度边和厚度边组成的面后，其余边和面分成上、下两组．两组的几何边界和拓扑结构是相同的，上、下两面的法向矢量方向相反．通过对读入的IGES三维线框文件进行预处理，使之满足特征识别的需要，采用基于几何推理的特征识别技术实现了对冲压件主型面特征、冲孔特征、翻孔特征的识别与提取，从而为后期冲压工艺方案的制定做准备．