Quantum dots trigger hot-start effects for pfu-based polymerase chain reaction

Hot-start (HS) effects were investigated in pfu-based polymerase chain reaction (PCR), when water-soluble CdTe quantum dots (QDs) were introduced in the PCR system. The HS effects were demonstrated by the higher amplicon yields and excellent suppression of non-specific amplification after pre-incubation of PCR mix with QDs between 35°C and 56°C. DNA targets were well amplified even after PCR mixture was pre-incubated 1?h at 50°C. Importantly, the effects of QDs nanoparticles could be reversed by increasing the pfu polymerase concentration, suggesting that there was an interaction between QDs and pfu DNA polymerase. Moreover, control experiment indicated that HS effect is not primarily due to the reduced pfu polymerase concentration resulted from the above interaction. Fluorescence correlation spectroscopy (FCS), a single molecule detection method, was used to investigate the possible mechanism of HS PCR with QDs. Preliminary FCS results suggested that CdTe QDs may directly interact with pfu DNA polymerase, rather than other components in the PCR system. Furthermore, results demonstrated that the interaction between QDs and pfu resulted in a reduction in pfu polymerase concentration. This study provided a good start to investigate potential implications of QDs in other key molecular biology techniques.     

Scalable arrays of chemical vapor sensors based on DNA-decorated graphene

Arrays of chemical vapor sensors based on graphene field effect transistors functionalized with single-stranded DNA have been demonstrated. Standard photolithographic processing was adapted for use on large-area graphene by including a metal protection layer, which protected the graphene from contamination and enabled fabrication of high quality field-effect transistors （GFETs）. Processed graphene devices had hole mobilities of 1,640 ± 250 cm2.V-1.s-1 and Dirac voltages of 15 ± 10 V under ambient conditions. Atomic force microscopy was used to verify that the graphene surface remained uncontaminated and therefore suitable for controlled chemical functionalization. Single-stranded DNA was chosen as the functionalization layer due to its affinity to a wide range of target molecules and π-π stacking interaction with graphene, which led to minimal degradation of device characteristics. The resulting sensor arrays showed analyte- and DNA sequence-dependent responses down to parts-per-billion concentrations. DNA/GFET sensors were able to differentiate among chemically similar analytes, including a series of carboxylic acids, and structural isomers of carboxylic acids and pinene. Evidence for the important role of electrostatic chemical gating was provided by the observation of understandable differences in the sensor response to two compounds that differed only by the replacement of a （deprotonating） hydroxyl group by a neutral methyl group. Finally, target analytes were detected without loss of sensitivity in a large background of a chemically similar, volatile compound. These results motivate further development of the DNA/graphene sensor family for use in an electronic olfaction system.     

基于隐马尔科夫模型的DNA序列分类方法

DNA序列分类是生物信息学的一项基础任务,目的是根据结构或功能的相似性预测DNA序列所属的类别。为进行有效分类,如何将序列映射到特征向量空间并最大程度地保留序列中蕴含的碱基间顺序关系是一项困难的任务。为克服现有方法容易导致因DNA序列碱基残缺而影响分类精度等问题,提出一种新的DNA序列特征表示方法。新方法首先为每条序列训练一个隐马尔科夫模型（HMM）,然后将DNA序列投影到由HMM状态转移概率矩阵的特征向量构成的向量空间中。基于这种新的特征表示法,构造了一种 K-NN分类器对DNA序列进行分类。实验结果表明,新型特征表示方法可以较为完整地保留 DNA 序列中不同碱基间的关系,充分反映序列的结构信息,从而有效提高了序列的分类精度。     

A rapid polymerisation process realizing 3D biocompatible structures in a microfluidic channel suitable for genetic analysis

Surface probe immobilisation is a complex and time consuming task undertaken prior to microfluidic integration, this requires surface functionalisation, biomolecule spotting, incubation and blocking steps. Traditional bonding techniques (anodic, thermal, etc.) or adhesives (UV cured) used to seal fluidic systems may denature biomolecules due to high temperature or vapour effects, thus bonding techniques such as thin film laminate or PDMS are used to seal systems, with substrate-fluidic alignment required prior to bonding. We propose a technique allowing probe DNA molecules to be immobilised in a sealed microfluidic system using (3D) hydrogel structures without any alignment steps. A prepolymer solution is introduced to the channels where photo-polymerisation is undertaken forming 3D structures covalently attached to the channel surface. We use a photo-initiated prepolymer material poly-ethylene-glycol (PEG) to form structures containing probe DNA. This process is fast compared to conventional biomolecule immobilisation techniques and is also biocompatible, this direct write approach removes overnight immobilisation/incubation of the probe DNA, it also facilitates immobilisation within a sealed fluidic system where conventionally DNA probe spots must be immobilised prior to channel sealing. We consider the transport of target DNA from bulk analyte to the 3D gel structure and evaluate hybridisation within the microfluidic system.     

基于DNA编码遗传算法构造广义决策树的研究

决策树是归纳学习和数据挖掘的重要方法,主要用于分类和预测。文章引入了广义决策树的概念,实现了分类规则集和决策树结构的统一。同时,提出一种新颖的基于DNA编码遗传算法构造决策树的方法。先用C4．5算法对数据集进行分类得到初始规则集,再通过文章中算法优化规则集并由此构建决策树。实验证明了该方法有效地避免了传统决策树构建过程的缺点,且有较好的并行性。     

基因表达数据中的局部模式挖掘研究综述

基因微阵列(DNA microarray)是实验分子生物学中的一个重要突破,其使得研究者可以同时监测多个基因在多个实验条件下表达水平的变化,进而为发现基因协同表达网络、研制药物、预防疾病等提供技术支持.研究者们提出了大量的聚类算法来分析基因表达数据,但是标准的聚类算法(单向聚类)只能发现少量的知识.因为基因不可能在所有实验条件下共表达,也不可能展示出相同的表达水平,但是可能参与多种遗传通路.在这种情况下,双聚类方法应运而生.这样就将基因表达数据的分析从整体模式转向局部模式,从而改变了只根据数据的全部对象或属性将数据聚类的局面.主要从局部模式的定义、局部模式类型与标准、局部模式的挖掘与查询等方面进行了梳理.介绍了基因表达数据中局部模式挖掘当前的研究现状与进展,详细总结了基于定量和定性的局部模式挖掘标准以及相关的挖掘系统,分析了存在的问题,并深入探讨了未来的研究方向.     

计算智能融合应用研究

本文对神经网络（NN），进化算法（EA0，不确定理论（Uncertain Theory）混沌系统（Chaos System）、免疫算法（IA,Immune Algorithm）、DNA计算（DNA Coputing）等广义计算智能相互融合应用进行了总结和分析，重点介绍免疫算法、DNA计算等新型计算智能理论的相互融合应用，最后对计算智能理论的融合应用趋势进行探讨。     

DNA分子计算模型

The field of practical DNA computing opened in 1994 with Adleman's paper,in which a laboratory experi-ment involving DNA molecules was used to solve a small instance of the Hamiltonian Path problem. The characteris-tic of this computation is its powerful ability in parallelism,its huge storage and high energy efficiency. This paper mainly introduces the principles of DNA computing and the sticker computing model.     

最小顶点覆盖问题的闭环DNA算法

提出了闭环DNA计算模型的基本概念及其基本生化实验,并给出了解决最小顶点覆盖问题的闭环DNA算法。在闭环DNA算法中,提出并实现了用删除实验直接构造顶点覆盖补集的构想;再通过电泳实验得到最小顶点覆盖的补集,由补集得到最小顶点覆盖。这使得算法的设计独特而新颖;由于算法仅用到基本的生化实验,这使得算法的实现简捷、可靠。