ABSTRACT

We outline a general methodology based on computational optimization and experimental data to reconstruct human pancreatic islet architectures. By using the nuclei coordinates of islet cells obtained through DAPI staining, cell types identified by immunostaining, and cell size distributions estimated from capacitance measurements, reconstructed islets composed of non-overlapping spherical cells were obtained through an iterative optimization procedure. In all cases, the reconstructed architectures included >99% of the experimental identified cells, each of them having a radius within the experimentally reported ranges. Given the wide use of mathematical modeling for the study of pancreatic cells, and recently, of cell-cell interactions within the pancreatic islets, the methodology here proposed, also capable of identifying cell-to-cell contacts, is aimed to provide with a framework for modeling and analyzing experimentally-based pancreatic islet architectures.

摘要

我们概述了一种基于计算优化和实验数据的通用方法，以重建人类胰岛结构。通过使用通过 DAPI 染色获得的胰岛细胞的细胞核坐标、通过免疫染色确定的细胞类型以及从电容测量估计的细胞大小分布，通过迭代优化程序获得由非重叠球形细胞组成的重建胰岛。在所有情况下，重建的架构包括 > 99% 的实验确定的细胞，每个细胞的半径都在实验报告的范围内。鉴于数学建模广泛用于研究胰腺细胞，以及最近胰岛内细胞间相互作用的广泛应用，这里提出的方法也能够识别细胞间接触，