Kinetic analysis of minimal residual disease using 3-dimentinal computational models of human granulopoiesis and development of risk of relapse-stratified treatment of pediatric leukemia.

2005年，我们试图开发一种独特的人类粒细胞生成计算模型，以确定稳态造血的调节机制。造血系统数学建模的计算方法已被应用于探索该系统的基本原理。模型需要易于访问，并且高度期望能够整合细节和再现细胞的动态行为，同时也可以预测响应各种模拟的特定细胞行为。我们利用三维细胞自动机（3D-CA）成功地开发了一种新的人体颗粒生成的横向计算模型，该模型结合了时空概念来描述在骨髓腔有限空间中发展的颗粒生成过程。我们强调该模型的独特性如下：1)该模型既不包含控制方程，也不包含用于调节的负反馈回路；2)人们可以直接在计算机屏幕上的硅颗粒生成中看到这一点，其中造血干细胞在由130 x 130 x 130单位立方面积组成的分析空间中复制、分化和分布它们的后代，其中结构对象假设为血管和小梁骨。该模型证实了单个颗粒细胞的局部相互作用产生的反馈回路导致系统的全局动态和稳定性的原理。2006年，我们进一步应用该模型分析了化疗和造血干细胞移植下颗粒生成的细胞动力学。模拟研究预测了G-CSF调度对药物性中性粒细胞减少的疗效和最低需求。归巢造血干细胞的细胞动力学解释了在不同干细胞来源的移植中观察到的移植期的差异。从计算血液学中学习细胞行为可以为血液系统恶性肿瘤的治疗提供新的策略。

In 2005, we attempted to develop a unique computational model of human granulopoiesis to identify the regulatory mechanisms for homeostatic hematopoiesis. A computational approach with mathematical-modeling of the hematopoietic system has been applied for exploring the principles that underlie the system. Models are required of easy access and highly expected to integrate details and reproduce the dynamic behaviors of the cells but also to predict the specific cellular behaviors in response to various simulations. We have successfully developed a new lateral computational modeling for human granulopoiesis using three-dimensional Cellular Automata (3D-CA), which incorporates a spatio-temporal concept to describe the granulopoietic process developing in the finite space of bone marrow cavity. We emphasize unique properties of this model as following ; 1)the model does not contain either governing equations or negative feedback loops assumed for regulation and 2)one can directly view this in silico granulopoiesis on the computer screen, in which HSCs replicate, differentiate, and distribute their offspring in an analytical space consisting of 130 x 130 x 130 unit cubic areas with structural objects assuming vessels and trabecular bones. The model reified the principle that local interaction of individual granulopoietic cells produced feedback circuits leading global dynamics and stability of the system.In 2006, we further applied this model to analyze cellular dynamics of granulopoiesis under chemotherapy and hematopoietic stem cell transplantation. Simulation studies predicted efficacy and minimal requirement of G-CSF scheduling for drug-induced neutropenia. Cellular dynamics of homing HSCs explained difference of engraftment periods observed in different utilization of stem cell sources in transplantation. Learning cellular behaviors from computational hematology could provide the novel strategies for the treatment of hematological malignancies.