基于pCTL的循环优化测试用例自动生成方法的研究与实现

编译优化是现代编译器的重要功能，编译优化测试对保障现代编译器质量有着重要作用。编译优化测试需要编写大量的测试用例程序作为输入，手工完成十分费时费力，因此，有必要研究编译优化测试用例的自动生成方法。 针对编译优化测试，有研究者提出并实现了一种基于分支时序逻辑的测试用例自动生成方法COT。该方法可以有效地生成标量优化测试用例程序，但该方法没有考虑程序中的数组和指针，对优化特征的描述也不能区分循环迭代中的语句实例，更不能刻画语句实例间的数据依赖关系，然而这些是刻画循环优化所必需的，因此，COT不适用于循环优化测试用例的自动生成。 本文针对COT方法的缺陷，首先提出了基于参数化分支时序逻辑pCTL (parameterized computation temporal logic)的循环优化描述方法，通过参数化COT的优化描述体系中的语句谓词、变量引用谓词和变量定义谓词实现了对循环迭代中的语句实例及语句实例间的数据依赖关系的描述。在此基础上，本文提出了基于pCTL 描述的循环优化测试用例自动生成方法COT2。该方法根据pCTL公式构造初始的关键节点控制图，再按照公式语义执行公式处理和虚边替换，得到完整的控制流图，最后计算数组引用下标，生成循环优化测试用例程序。 本文实现了COT2的系统原型，并用循环优化模块的覆盖率指标评价生成的测试用例的质量，实验结果表明该方法对循环优化具有针对性，是一种行之有效的方法。本文还用生成的测试用例程序对GCC各版本的循环优化模块进行了测试，并分析了错误发现数与各版本稳定性之间的关系，进一步验证了本文方法的有效性。

吲哚生物碱Desbromoarborescidine A 的全合成及烯胺C-亚磺酰化反应研究

从新几内亚核桃木的树皮中分离得到的吲哚类喹诺里西定生物碱10-Desbromoarborescidine A，因发现其具有阻滞钙离子通道的活性而倍受关注。10-Desbromoarborescidine A由A、B、C、D四个环组成，只有一个手性中心，是吲哚生物碱中结构较简单的一种，常作为此类生物碱全合成方法的模型化合物。但迄今为止，能高效而简便的实现手性10-Desbromoarborescidine A不对称全合成方法线路不多，大多数以不对称诱导的方式建立其手性中心，手性催化的方式仅有一例金属催化。从逆合成分析可知，Desbromoarborescidine A的全合成可以通过亚胺不对称催化还原进行关键的手性中心构建，而本课题组在之前的研究中通过手性有机小分子催化剂的发展，已将三氯硅烷氢转移还原亚胺发展成了一类简便实用、高效、高对映选择性并具有优良底物适应范围的不对称催化反应，我们希望以这一反应作为关键手段，发展一条Desbromoarborescidine A及其类似物不对称合成新路线。 根据我们设计的新路线，首先成功合成了其关键中间体，然后我们进行了关键的不对称催化尝试。用本实验室已有的高性能有机小分子催化剂虽得到了较好的对应选择性，但是产率很低。同时，为了验证整条线路的可行性，我们也用消旋的中间体进行拉通线路的尝试。但不幸的是，在脱除保护基时遇到了很大困难。尝试换不同的保护基，或改变脱保护基的顺序，都未能成功合成目标产物。究其原因可能是由于吲哚的特殊性造成的，吲哚类亚胺与常规的芳香亚胺有较大的差异，其NH基团无论保护还是不保护，对与其2位相联接的C=N双键均有很大的影响，导致其不对称催化还原难以进行。另外，由于所设计的还原产物含有处在吲哚苄位的胺基，稳定性较差，造成保护基脱除困难。 烯胺C-亚磺酰化反应是本课题组最近发现的一个新反应，之前未见文献报道。本研究对该反应进行了反应条件优化和底物扩展，发现带Cbz，Ac，COt-Bu，CO2Et，Bz等保护基的一系列环状和非环状烯胺在亚磺酸钠、DMAc和MeSiCl3的共同作用下能高效高产率生成β-胺基烯基亚砜类新化合物，为合成多官能团化的烯基亚砜新化合物提供了一条简便实用的途径。 The main constituent of Dracontomelum mangiferum B1, indoloquinolizidine alkaloid 10-Desbromoarborescidine A, has drawn great attention due to its calcium channel blocking activity. Its molecular structure is relatively simple compared with the other alkaloids of the same type, which has only one chiral center, albeit with four cycles A, B, C, and D. This compound is often used as a model target for exploring different strategies for the total synthesis of indole alkaloids. Nevertheless, so far there still lack practical and highly efficient methods for the asymmetric total synthesis of 10-Desbromoarborescidine A. Most of the current available methods rely on stoichiometric asymmetric synthesis for the construction of the chiral center. There is only one example reporting utilization of asymmetric catalysis, but with transition metal complex as the catalyst. Our retrosynthetic analysis shows that catalytic asymmetric reduction of imine could be used as the key step for the construction of the chiral center of Desbromoarborescidine A. Since in the previous studies our group has developed the asymmetric reduction of imines by trichlorosilane into a practical and highly efficient and enantioselective method using newly designed chiral organocatalysts, we hope to apply this method to develop a novel synthetic route for the total synthesis of Desbromoarborescidine A and its analogues in this study. According to the newly designed synthetic route, we first accomplished the synthesis of the key intermediates which was then examined for the critical asymmetric catalysis. The asymmetric reduction using the highly efficient organocatalysts, developed in our lab afforded high ee but poor yield. We tried different reaction conditions to improve the yield, but failed to get any good results. Simultaneously, to vertify the feasibility of the synthetic route we designed, we also tired to go through the route toward the racemic synthesis of Desbromoarborescidine A. But unfortunately, protection and deprotection proved to be big hurdles. All the different protection groups and different sequences of protection and deprotection we tried failed to get us through the designed synthetic sequence and furnish the final product. Most likely, the indole part is the culprit behind the failures.The NH group of the indole, no matter protected or not, may impact the catalytic asymmetric reduction of C-N double bond connected with 2-C. Additionally, the reduction product we designed contains an amino group in the β-position of the indole, which may cause problems due to its instability. C-sulfenylation of enamines is a novel reaction discovered recently by our group, which has not been seen before in the literature. In this study, optimization of the reaction conditions and exploration of the substrate scope were further undertaken for this reaction, which reveal that a series of enamines with N-Cbz, Ac, COt-Bu, CO2Et protection groups could all undergo smooth C-sulfinylations with the comined use of sodium benzene sulphinate, DAMc and MeSiCl3, efficiently furnishing the β-amino vinylsulfoxide products in high yield, affording a practical and highly efficient methods for synthesis of functional vinylsulfoxides.

SYNTHESES AND CRYSTAL-STRUCTURES OF (C8H8)ND(C5H9C5H4)(THF)(2) AND [(C8H8)GD(C5H9C4H4(THF)][(C8H8)GD(C5H9C5H4(THF)(2)]

LnCl(3) (Ln = Nd, Gd) reacts with C5H9C5H4Na (or K2C8H8) in THF (C5H9C5H4 = cyclopentylcyclopentadienyl) in the ratio of 1:1 to give (C5H9C5H4)LnCl(2)(THF)(n) (or (C8H8)LnCl(2)(THF)(n)], which further reacts with K2C8H8 (or C5H9C5H4Na) in THF to form the title complexes. If Ln = Nd the complex (C8H8)Nd(C5H9C5H4)(THF)(2) (a) was obtained: when Ln = Gd the 1:1 complex [(C8H8)Gd(C5H9C5H4)(THF)][(C5H8)Gd(C5H9C5H4)(THF)(2)] (b) was obtained in crystalline form. The crystal structure analysis shows that in (C8H8)Ln(C5H9C5H4)(THF)(2) (Ln = Nd or Gd), the Cyclopentylcyclopentadienyl (eta(5)), cyclooctatetraenyl (eta(8)) and two oxygen atoms from THF are coordinated to Nd3+ (or Gd3+) with coordination number 10. The centroid of the cyclopentadienyl ring (Cp') in C5H9C5H4 group, cyclooctatetraenyl centroid (COT) and two oxygens (THF) form a twisted tetrahedron around Nd3+ (or Gd3+). In (C8H8)Gd(C5H9C5H4)(THF), the cyclopentyl-cyclopentadienyl (eta(5)), cyclooctatetraenyl (eta(8)) and one oxygen atom are coordinated to Gd3+ with the coordination number of 9 and Cp', COT and oxygen atom form a triangular plane around Gd3+, which is almost in the plane (dev. - 0.0144 Angstrom).

牡蛎染色体的分子生物学分析

本研究应用显带技术和荧光原位杂交(Fluorescence in situ hybridization，FISH)技术，鉴定了牡蛎的染色体；应用FISH方法定位了一系列的重复序列和大分子的P1克隆DNA；制备了染色体特异性探针。应用FISH特异性探针成功地鉴定了长牡蛎的三体10。结果如下：1．分析了G带和C带在美洲牡蛎染色体上的分布。G带在每一条染色体上的带型不同，某些染色体间(如第1对和第4对染色体，第7对和第9对染色体)的带型差别不是很明显。G带型容易受染色体收缩程度的影响。C带型重复性较好，染色体带型较清楚，分布在染色体的端粒区域和着丝粒区域。G带和C带带型能够用来鉴定牡蛎的染色体，但是重复性低和带型差异不显著，并不适合常规的染色体鉴定。2．早期胚胎和担轮幼虫制备的染色体适合于FISH分析。染色体制备方法重复性好，可适用于其它贝类的染色体制备。3．研究了重复序列基因--rDNA的定位：1)18S-5.8S rDNA在研究的五种巨蛎属Crassostrea牡蛎均只有一个位 点。太平洋种(C.gigas，C. ariakensis和C. plicatula)中，杂交信号位于最短的染色体一第10对染色体长臂的端粒区域，在大西洋种(C. virginica和C. rhizophorae)中，同一序列定位在第2对染色体短臂的端粒区域。2)18S-28S rDNA在两种蛤中有两个位点。rDNA探针定位在侏儒蛤(Mulinis Lateralis)的第15对和第19对染色体的端粒区域，同一序列定位在硬壳蛤(Mercenaria mercenaria)的第10对染色体的长臂和第12对染色体短臂的端粒区域。信号强度在两对染色体之间有差异。 3)5s rDNA位于美洲牡蛎的第5对染色体的短臂上靠近着丝粒区域和第6 对染色体的短臂的中间区域。信号强度在两对染色体之间没有显著差异。5S rDNA探针可以作为鉴定和识别第5对和第6对染色体的特异性探针。4．研究了一些重复序列的定位1)两个短的重复序列1G8，1P2均产生很强的荧光信号分布在美洲牡蛎所有的染色体上。在低严谨条件下，这些序列均产生很强的信号散布在所有的染色体上。在高严谨条件下，信号强度大大减弱，但是信号仍散布在所有的染色体上。这些重复序列散布在美洲牡蛎的整个基因组中。2)高度重复序列Cgl70产生的信号分布在长牡蛎的7对染色体的着丝粒区域，没有发现间区信号。在第1对，第2对，第4对和第7对染色体上的荧光信号强且稳定。在第5对，第8对和第10对染色体上的信号相对弱且不稳定。在剩余的染色体上(第3对，第6对和第9对染色体)没有检测到荧光信号。结果表明此卫星序列是一个着丝粒卫星序列。在美洲牡蛎的染色体上没有检测到荧光信号，表明了这个着丝粒卫星序列在这两种牡蛎中的分布存在着显著的差异。3)脊椎动物端粒序列(TTAGGG)n的FISH信号局限在四种双壳贝类(美洲牡蛎，the mangrove oyster，硬壳蛤，侏儒蛤)所有染色体的端粒区域，没有发现间区信号的存在。研究结果与已报道的研究结果表明脊椎动物端粒序列或许存在于所有双壳贝类的染色体末端。双壳贝类是目前研究过的唯一含有脊椎动物端粒序列DNA的无脊椎动物。4)研究了RAPD探针在美洲牡蛎染色体上的定位。大多数RAPD探针产生了多个信号散布在间期细胞核和所有的染色体上。引物OPX-03，OPX-04，OPX—06，OPG-02，OPM—04，OPM-11，0PS-02制备的探针在适宜的条件下产生特异性荧光 信号，分布在牡蛎的特定的染色体上。PCR特异性带产生的探针OPX—06—310和0PG-02—300产生了特异性的荧光信号：OPX—06—310产生的信号位于第5对染色体的短臂的近端粒区域，0PG—02—300探针定位到第3对染色体的短臂上。这两个探针是鉴定美洲牡蛎单条染色体的特异性探针。5．研究了大分子Pl克隆DNA(插入片断为80～100 kb)在美洲牡蛎染色体上的定位。Pl克隆DNA通过切口平移方法标记digoxigenin—11-dUTP用作FISH的探针。Cot-1 DNA作为竞争剂有效地抑制了Pl克隆序列中的重复序列产生的信号。杂交信号用fluorescein标记的anti—digoxigenin抗体来检测，用两层抗体rabbit-anti-sheep抗体和FITC anti—rabbit抗体来扩增信号。9个P1探针成功地定位在特定的染色体上。46—1探针杂交到第1对染色体的长臂靠近着丝粒区域；47-10探针定位到第2对染色体的长臂近端粒区域；Cvpl和48-13两探针定位到第3对染色体上：Cvpl位于短臂的端粒区域，48-13探针位于长臂的近着丝粒区域；48—10探针杂交到第4对染色体的长臂上；48-1探针杂交到第5对染色体长臂的近着丝粒区域；49-11探针位于第7对染色体长臂上；探针49-10和44-11位于第8对染色体长臂上。同时我们成功地将2个P1探针杂交到同一染色体分裂相中，进一步确定了Pl探针在美洲牡蛎染色体 上的定位。6．应用18S-28S rDNA探针成功地鉴定出长牡蛎非整倍体中的三体10。经鉴定AF-35，AF-39和AF-3三体家系属于三体10家系。rDNA探针分布在三条染色体上，即多出的一条染色体为染色体10。相应地在间期细胞核上有三个信号出现。AF-34和AF-36家系不属于三体10家系。rDNA探针分布在两条染色体上，相应地在间期细胞核上有两个信号出现。FISH和染色体特异性探针为非整倍体的鉴定提供了一个快速准确可靠的方法和途径。

Characterization of eastern oyster (Crassostrea virginica Gmelin) chromosomes by fluorescence in situ hybridization with bacteriophage P1 clones

Chromosome identification is an essential step in genomic research, which so far has not been possible in oysters. We tested bacteriophage P1 clones for chromosomal identification in the eastern oyster Crassostrea virginica, using fluorescence in situ hybridization (FISH). P1 clones were labeled with digoxigenin-11-dUTP using nick translation. Hybridization was detected with fluorescein-isothiocyanate-labeled anti-digoxigenin antibodies and amplified with 2 layers of antibodies. Nine of the 21 P1 clones tested produced clear and consistent FISH signals when Cot-1 DNA was used as a blocking agent against repetitive sequences. Karyotypic analysis and cohybridization positively assigned the 9 P1 clones to 7 chromosomes. The remaining 3 chromosomes can be separated by size and arm ratio. Five of the 9 P1 clones were sequenced at both ends, providing sequence-tagged sites that can be used to integrate linkage and cytogenetic maps. One sequence is part of the bone morphogenetic protein type 1b receptor, a member of the transforming growth factor superfamily, and mapped to the telomeric region of the long arm of chromosome 2. This study shows that large-insert clones such as P1 are useful as chromosome-specific FISH probes and for gene mapping in oysters.