Development of computational tools for studying protein sequences, structures and signaling networks

细胞由蛋白质信号网络组成，这些网络执行生物学功能，例如调节细胞生长或催化生化反应。 因此，蛋白质的功能障碍往往会导致人类疾病，如阿尔茨海默氏病，心脏病和癌症。 我的长期研究目标是创造合成蛋白质信号网络，使我们有一天能够像电路和计算机网络一样精确地操纵细胞生物学。 为了实现这一目标，我的建议将集中于开发用于研究蛋白质序列、结构和信号网络的计算工具。 首先，为了推断蛋白质序列的功能，使用Smith Waterman（SW）算法来找到其与已知功能的蛋白质的相似性。 随着序列数据库变得越来越大，需要更快的序列比较方法，例如使用加速的现场可编程门阵列（FPGA）硬件。 为了使FPGA解决方案更经济实惠，我将开发FPGA硬件，以使用更少的资源加速SW算法，同时保持相当的速度。 接下来，为了研究细胞内的蛋白质信号动力学，荧光蛋白生物传感器是强大的工具，但这些生物传感器的设计往往是反复试验。 使用计算工具来模拟蛋白质生物传感器的构象空间，我改进了设计，但该工具不是定量的。 为了解决这个问题，我将包括选择优选的生物传感器构象的分子因素。 最后，为了设计合成蛋白质网络或对更大的现有网络建模，我将开发一个计算工具来模拟蛋白质信号网络的时空动力学。 这项工作将有助于深入了解蛋白质序列及其网络，最终有助于开发人类疾病的治疗方法。

Cells are composed of protein signaling networks that perform biological functions such as regulating cell growth or catalyzing biochemical reactions. As a result, the malfunction of proteins often causes human illnesses such as Alzheimer's disease, heart disease and cancer. My long term research goal is to create synthetic protein signaling networks that will allow us to one day manipulate cell biology with the same precision as electrical circuits and computer networks. To accomplish this goal, my proposal will focus on developing computational tools for studying protein sequences, structures and signaling networks. First, to infer the function of a protein sequence, the Smith Waterman (SW) algorithm is used to find its similarity to proteins of known function. As sequence databases grow larger, faster sequence comparison approaches are required such as using accelerated field programmable gate array (FPGA) hardware. To make the FPGA solution more affordable, I will develop FPGA hardware for accelerating the SW algorithm using fewer resources while maintaining a comparable speed. Next, to study the protein signaling kinetics within cells, fluorescent protein biosensors are powerful tools but the design of these biosensors is often trial and error. Using a computational tool to model the conformational space of protein biosensors, I improved the design however the tool was not quantitative. To address that problem, I will include molecular factors that select preferred biosensor conformations. Lastly, to design synthetic protein networks or model larger existing networks, I will develop a computational tool for simulating the spatial and temporal kinetics of protein signaling networks. Together this work will yield insights into protein sequences and their networks that will ultimately aid in developing therapies for human illnesses.