Center

斯坦福癌症系统生物学中心(CCSB)旨在发现分子机制 通过将癌症作为一个复杂的生物系统来研究癌症的进展，该系统在一定程度上是由 分化受损。越来越多的证据表明，许多癌症像正常组织一样，是由 处于不同分化阶段的细胞层级，疾病通过自我更新来维持 亚群。我们的首要目标是提供对自我更新特性的更好理解 癌症将使我们能够确定根除这种疾病的分子治疗靶点和策略，或者 使其处于非致命性状态。我们的生物项目与新的计算技术相结合， 旨在剖析作为癌症驱动因素的分化受损的过程和原因 几种血液系统恶性肿瘤的进展。这种方法将使我们能够确定差异。 这些恶性肿瘤和其他癌症的共性之间的关系。 为了确定癌症进展的机制基础，基于网络的多尺度 视点是必填项。癌症等疾病越来越多地被认为是由于 复杂生物系统的协调性能。这种系统生物学观点使得 结合高吞吐量、高维数据和计算方法的发展 特别是针对它的分析。有三个基本和内部锁定要求，以全面 癌症的系统分析。首先，需要强大的方法来推断分子调控网络 驱动分化等表型过程。其次，计算方法需要能够 识别和分离癌症的潜在进展模式，然后将其与潜在的 监管网络。第三，可执行的模型是可取的，这样就有可能提出假设性的“假设” 例如，预测有针对性的干预可能如何影响后续病程的问题。 我们将作为CCSB开发的方法针对这三个具体的计算目标。它们是量身定做的 为了解决我们正在研究的生物系统，我们的总体CCSB目标是了解 分化和自我更新癌症。然而，它们将具有更广泛的适用性。因此，虽然在这里 我们将它们应用于特定的生物系统，模型预测的实验测试将不仅验证 生物学结论，还有方法论本身。此外，实验验证将 在反复完善和改进我们的计算模型方面发挥关键作用。 恶性血液病为研究自我更新和分化的作用提供了独特的机会 在癌症中。免疫系统的细胞由造血干细胞(HSCs)通过分层过程发展而来 分化成更特殊的细胞类型，这一点已经被很好地定义和研究。自我更新的HSC 最初产生具有分化为多细胞潜能的多能祖细胞(MPP) 类型，但缺乏自我更新能力。肿瘤中的MPP可产生低电位型共同髓系祖细胞(CMP) 和共同淋巴祖细胞(CLP)，生成主要的髓系和淋巴系，包括 免疫系统。随后的分化会逐渐产生更专业的细胞类型，这些细胞缺乏 自我更新能力，最终导致主要效应细胞，如T细胞，B细胞，巨噬细胞和 粒细胞。我们将剖析导致放松管制的分化和获得异常的过程 髓系和淋巴系的自我更新能力。为此，我们将调查三个 补充系统：人类急性髓系白血病(AML)、人类滤泡性淋巴瘤(FL)和 人和小鼠T细胞急性淋巴细胞淋巴瘤(T-ALL)。 我们的计算方法产生了分子和细胞相互作用的网络级表示 集成多个尺度上的各种数据类型(分子、细胞表型、肿瘤表型、 临床结果)，并通过分化和自我更新途径的观点筛选结果。通过 结合实验和计算方法，我们的目标是预测和验证临界像差 建立和维持癌症自我更新能力的分子事件及其与 正常细胞层级中的分化。我们的方法是基于机器学习的，可执行 几何和拓扑学中的模型、多尺度建模和方法。将会有一个 与实验项目密切互动，在迭代过程中计算的生物验证 预测为改进计算模型提供了基础。出于这个原因，计算方法 开发将在一个项目下进行，该项目将与我们CCSB的所有实验小组密切互动。 斯坦福CCSB代表着我们作为U56 ICBP规划中心的现状的演变。在我们的 跨物种系统生物学分析FL转化和转基因小鼠模型的作用 差异化(特别是自我更新计划的异常激活)成为统一的关键 癌症进展的主题。这项建议是以我们的发现为基础的。我们将扩展我们的集成系统 研究分化和自我更新在癌症中的作用，以及正常的调控网络如何 在癌症中，控制这些过程变得不受管制。

The Stanford Center for Systems Biology of Cancer (CCSB) aims to discover molecular mechanisms underiying cancer progression by studying cancer as a complex biological system that is driven, in part, by impaired differentiation. Increasing evidence indicates that many cancers, like normal tissue, are composed of a hierarchy of cells at different stages of differentiation, and that the disease is maintained hy a self-renewing subpopulation. Our overarching goal is to provide a better understanding of the self-renewing properties of cancer that will enable us to identify molecular therapeutic targets and strategies to eradicate this disease, or to maintain it in a nonlethal state. Our biological projects are integrated with novel computational techniques, designed to dissect processes and causal factors underlying impaired differentiation as a driver of cancer progression in several hematologic malignancies. This approach will enable us to ascertain differences between these malignancies, and commonalities which may generalize to other cancers. In order to identify mechanistic underpinnings of cancer progression, a network-based and multiscale viewpoint is mandatory. Increasingly, diseases such as cancer are recognized as resulting from disruption in the coordinated performance of a complex biological system. This systems biology viewpoint necessitates the incorporation of high throughput, high dimensional data, and development of computational methods specifically geared to its analysis. There are three essential and interiocking requirements for a comprehensive systems analysis of cancer. First, powerful methods are required to infer molecular regulatory networks that drive phenotypic processes such as differentiation. Second, computational approaches are needed that can identify and isolate underlying patterns of progression in cancer, which can then be related to underlying regulatory networks. Third, executable models are desirable so that it is possible to pose hypothetical "what if' questions to predict how, for example, a targeted intervention might affect the subsequent course of disease. The approaches we will develop as a CCSB target these three specific computational aims. They are tailored to address the biological systems we are studying in our overall CCSB goal to understand the role of differentiation and self-renewal cancer. However, they will have much wider applicability. Thus, although here we apply them to particular biological systems, experimental testing of model predictions will validate not only the biological conclusions, but also the methodologies themselves. Furthermore, experimental validation will play a crucial role in iteratively refining and improving our computational models. Hematologic malignancies provide a unique opportunity to study the role of self-renewal and differentiation in cancer. Cells ofthe immune system develop from hematopoietic stem cells (HSCs) by a hierarchical process of differentiation to more specialized cell types, that has been well defined and studied. Self-renewing HSCs give rise initially to multipotent progenitors (MPPs) that have the potential to differentiate into multiple cell types, but lack self-renewal capacity. MPPs in tum give rise to oligopotent Common Myeloid Progenitor (CMP) and Common Lymphoid Progenitor (CLP), generating the major myeloid and lymphoid lineages that comprise the immune system. Subsequent differentiation produces progressively more specialized cell types that lack self-renewal ability, ultimately resulting in the major effector cells such as T-cells, B-cells, macrophages, and granulocytes. We will dissect the processes leading to deregulated differentiation, and acquisition of aberrant self-renewal ability in both myeloid and lymphoid lineages. For this purpose we will investigate three complementary systems: human Acute Myeloid Leukemia (AML), human Follicular Lymphoma (FL), and human and mouse T-cell Acute Lymphoblastic Lymphoma (T-ALL). Our computational methods produce network-level representations of molecular and cellular interactions that integrate diverse data types across multiple scales (molecular, cellular phenotypes, tumor phenotype, clinical outcomes) and filter the results through the viewpoint of differentiation and self-renewal pathways. By combining experimental and computational methods, we aim to predict and validate the critical aberrant molecular events that establish and maintain the self-renewal capacity of cancer, and how they relate to differentiation in normal cellular hierarchies. Our approaches are based on machine learning, executable models, multiscale modeling, and methods from the mathematics of geometry and topology. There will be a close interaction with experimental projects, in an iterative process where biological validation of computational predictions provides the basis for improved computational models. For this reason, computational methods development will occur under one project that interacts closely with all the experimental groups in our CCSB. The Stanford CCSB represents an evolution from our current status as a U56 ICBP Planning Center. In our cross-species systems biology analysis FL transformation and transgenic mouse models, the role of differentiation (and particulariy the aberrant activation of self-renewal programs) emerged as a key unifying theme in cancer progression. This proposal builds on our findings. We will extend our integrated systems studies into the role of differentiation and self-renewal in cancer, and how normal regulatory networks governing these processes become deregulated in cancer.