Voxelation: a new method for 3D gene expression analysis

描述(申请人提供)：后基因组时代的重大机遇 世界将会理解基因组是如何构造大脑的，以及这是如何 过程在疾病中出了差错。我们采取的方法结合了 生物医学成像技术的计算过程，如CT和 PET，用高通量的方法获得三维基因表达模式 大脑。这项新技术被称为体素释放，它的名字来源于 术语“体素”，指的是立方3D图像体元素。体素是 3D模拟熟悉的2D图像元素，即像素。Voxelation需要 在大脑的空间记录中直接创建体素(立方体)，以及 高通量基因表达分析技术在RNA中的应用 从体素中提取。然后使用成像算法来重建 空间信息的高通量。利用微阵列，我们已经获得了较低的 正常小鼠大脑的9000个基因的分辨率(40体素)图像，给出了 对基因组构建大脑的新见解。还在使用 基因芯片，我们已经获得了2000个基因，24个体素的冠状图像 正常和阿尔茨海默病人脑水平的半侧脑区 海马体的。我们的分析揭示了共同监管的共同网络 基因，并允许识别可能的控制区。此外，a 已经发现了一些有趣的候选基因，它们的表达水平 在阿尔茨海默氏症和正常的偏瘫之间有很大的不同。 最后，奇异值分解(SVD)，一种用于 简明扼要地解释复杂的数据集、生成的图像 区分大脑结构，包括皮质、尾状核和 海马体。在这些研究提供的基础上，体素处理 会用鼠标来实现两个特定的目标。(1.)利用微阵列， 大脑的低分辨率(40体素)高通量(9,000)基因图像将 在人类神经精神障碍的四个小鼠模型中获得： 帕金森病、药物成瘾、精神分裂症和多巴胺D3受体 缺乏症。(2)为了扩展体素处理技术，高分辨率(300 体素)中等吞吐量(数百个基因)的小鼠大脑图像将 采用实时荧光定量RT-PCR方法获得。一种名为体素加速器的仪器具有 是为体素的半自动小型化采集而建造的。 来自特定目标1和2的图像将使用经典图像进行验证 技术，如原位杂交。这项研究有望带来重要的新成果。 对基因组和大脑在健康和疾病方面的联系的洞察，并可能 也为治疗提供了新的途径。

DESCRIPTION (provided by applicant): A major opportunity for the post-genomic world will be to understand how the genome constructs the brain, and how this process goes awry in disease. The approach we are taking combines the computational procedures of biomedical imaging technologies, such as CT and PET, with high throughput methods to acquire 3D gene expression patterns in the brain. This new technology is called voxelation, and gets its name from the term "voxel," which refers to a cubic 3D image volume element. The voxel is the 3D analog of the familiar 2D image element, the pixel. Voxelation entails the direct creation of voxels (cubes) in spatial register with the brain, and application of high throughput gene expression analytic techniques to RNA extracted from the voxels. Imaging algorithms are then employed to reconstruct spatial information in high throughput. Using microarrays, we have obtained low resolution (40 voxel) images of the normal mouse brain for 9,000 genes, giving novel insights into the construction of the brain by the genome. Also using microarrays, we have acquired 2,000 gene, 24 voxel images of coronal hemisections of both normal and Alzheimer's disease human brains at the level of the hippocampus. Our analysis has revealed a common network of co-regulated genes, and allowed identification of putative control regions. In addition, a number of intriguing candidate genes have been found whose level of expression is significantly different between the Alzheimer's and normal hemisections. Finally, singular value decomposition (SVD), a mathematical technique used to provide parsimonious explanations of complex data sets, produced images that distinguished between brain structures, including cortex, caudate and hippocampus. Building on the foundations provided by these studies, voxelation will be employed in two Specific Aims using the mouse. (1.) Using microarrays, low resolution (40 voxel) high throughput (9,000) gene images of the brain will be acquired in four mouse models of human neuropsychiatric disorders: Parkinson's disease, drug addiction, schizophrenia, and dopamine D3 receptor deficiency. (2.) To extend the technology of voxelation, high resolution (300 voxel) moderate throughput (hundreds of gene) images of the mouse brain will be obtained using real-time quantitative RT-PCR. An instrument, the voxelator, has been constructed for semiautomated miniaturized harvesting of the voxels. Images from both Specific Aims 1 and 2 will be validated using classical techniques, such as in situ hybridization. This research promises important new insights into the nexus of genome and brain, in health and disease, and may also provide novel avenues to therapy.