Antisense Oligonucleotides targeting APP to prevent neurodegeneration in models of Down Syndrome and Alzheimer's disease

点击翻译按钮获取中文摘要

Project Summary/Abstract: Critical to ensuring rigor in investigation of neurodegenerative disorders is the use of methods by which to accurately, and in an unbiased way to, assess degenerative phenotypes, including the number of neurons in vulnerable populations. Unbiased stereological assessment is the standard in the field. However, these studies are technically demanding and are costly in terms of the equipment required to conduct them. Thus, limitation of access to existing, specialized and prohibitively expensive systems has been a hindrance to progress, especially during the Covid-19 pandemic. To overcome these limitations we propose to develop a novel scalable open-source stereological software system with drastically reduced system requirements and at the fraction of the cost of commercial stereology systems. The proposed system will be capable of running on standard windows systems and standard microscopy hardware that is widely available in laboratories. Increasing the availability of microscope systems capable of stereological analysis promises to dramatically improve access to stereological data collection, speed analyses and enhance rigor as well as transparency. As such, our goal is to achieve stereological protocol requirements with a standard lab microscope without the need for expensive specialized hardware. We will acquire statistically equivalent stereological data using innovative analytical components using well-established computer vision methods like “visual odometry” and “depth from focus” that will make use of camera images in combination with open-loop stage positioning data to replace the functionality of expensive feedback-based high precision 3D positioning sensors. An additional benefit to developing an open-source system architecture that de-couples hardware from software protocols is that it will more naturally allow integration of AI/machine learning based components to automate and speed manual stereology tasks like feature/cell detection and counting. To demonstrate reproducibility and statistical equivalence with currently in use commercial systems, we will use samples previously analyzed using existing systems. Future directions would include automated cell detection to improve the reproducibility and effeciency of performing stereology experiments.