植物线虫rDNA内转录间隔区亚太赫兹特征光谱表征研究

理论与实验研究表明，在亚太赫兹频段，标记生物大分子特征的DNA靶标序列呈现出许多由氢键、范德华力、偶极子的旋转和振动跃迁以及晶格和骨架的低频振动产生的微弱声子模式，它们与DNA的构成、极化方式、拓扑结构、生存环境及形态密切相关，可用作分子生物鉴别的光学探针。与传统的分子生物学检测方法相比具有无标记、无污染、快速、简便等特点。基于此，项目组开展了植物线虫rDNA亚太赫兹光谱的识别与表征研究，搭建低温探测系统，研究靶标序列及DNA结构单元与结晶水、薄膜水和结合水作用后的亚太赫兹光谱，深度理解水对线虫DNA光谱的影响和作用机制；同时利用数据挖掘技术、蒙特卡罗随机抽样与统计分析方法获得线虫rDNA特征序列；进而将分子动力学理论和周期性边界条件下基于能量的分块方法（PBC-GEBF）量子力学算法相结合，并辅助主成分分析和神经网络等机器学习算法对特征光谱进行分类和识别，以达到准确鉴别线虫的目标。

Theoretical and experimental studies show that in Sub-THz band, DNA target sequences that mark biological macromolecule features present many weak phonon modes caused by hydrogen bonds, van der Waals forces, dipole rotation and vibration transitions, and low-frequency vibration of the lattice and skeleton, which are closely related to the composition of DNA, polarization mode, topological structure, living environment and morphology, can be employed as optical probes for molecular biological identification. Compared with the traditional methods of molecular biology detection, it is no mark, no pollution, fast and simple. Based on this, the project team carried out the research on the identification and characterization of the plant nematode rDNA in Sub-THz region to study the spectra of target sequences and DNA structural units interacting with crystal water, film water and binding water, by constructing cryogenic detection system, to understand the influence and mechanism of water on the spectrum of nematode rDNA. Furthermore, data mining techniques and Monte-Carlo random sampling and statistical analysis methods were used to acquire the nematode rDNA signature sequences. Then the molecular dynamics theory and the generalized energy-based fragmentation approach under periodic boundary conditions (denoted as PBC-GEBF) methods were combined, together with the machine learning algorithms such as principal component analysis and neural network were applied to classify and identify the characteristic spectra, and finally to achieve the goal of accurately identifying nematodes.