Knowledge Discovery Applications in High-Throughput Polymer Characterization
- PDF / 282,575 Bytes
- 6 Pages / 612 x 792 pts (letter) Page_size
- 46 Downloads / 168 Views
0894-LL05-04.1
Knowledge Discovery Applications in High-Throughput Polymer Characterization Jing Su1, Pedro Zapata, and J. Carson Meredith School of Chemical & Biomolecular Engineering 1 Coulter School of Biomedical Engineering Georgia Institute of Technology Atlanta, GA 30332-0100 ABSTRACT The use of combinatorial libraries for screening cell-material interactions presents a significant need for robust informatics methods capable of extracting knowledge from large, combinatorial datasets. We describe the development of local quantitative metrics based on cellcell and cell-microstructure distributions about each individual cell. The local metrics are shown to detect sensitive effects of surface features on proliferation of MC3T3-E1 osteoblasts in situations where global summary statistics were unsuccessful. INTRODUCTION Cell-surface interactions are critical in life science, tissue engineering and medicine. As a significant surface feature, surface lateral patterns have crucial effects on cell behaviors as well as on tissue functions[1,2]. Extensive investigations show that cells may use different mechanisms to recognize lateral patterns of different size scales[3,4]. However, cell – surface lateral pattern interactions are still far from being well understood. We hypothesize that one difficulty in studying cell-surface interactions is the global measures of cell behavior and surface features that are often employed, whereas cell-recognizable environments are actually localized in a microscopic scale around each individual cell. This is believed to be the case in particular for large, combinatorial data sets. To address these challenges, in this study the local ‘point of view’ of individual cells was used to define mathematical descriptions of cell-cell and cell-surface interactions. These were then compared to global summary statistics in their effectiveness in mining large combinatorial data sets. METHODOLOGY AND EXPERIMENTAL DETAILS Combinatorial Library Preparation Poly-DL-lactic-glycolic acid (PLGA, block copolymer, 50:50 ratio of PGA and PLA; Mw=40,000~75,000Da) and poly-ε-capro-lactone (PCL Mw = 114,000Da, Mw/Mn = 1.43) were obtained from Sigma Aldrich, St Louis, MO. The PLGA/PCL composition-annealing temperature (Φ/T) two-dimensional libraries were prepared on 25mm×25mm silicon chips treated with APTES (3-amino-propyl-tri-ethoxy-silane, Sigma Aldrich, St Louis, MO). Annealing temperature (T, 93~130°C) and PCL composition (ΦPCL, 0~0.8, mass fraction) gradients were generated along orthogonal directions using procedures described previously. [5] The ΦPCL-gradient films were annealed on a custom aluminum T-gradient heating stage for 2 h, with the T gradient orthogonal to the ΦPCL-gradient, and were immediately quenched to room temperature[5]. Driven by crystallization and LCST (lower critical solution temperature) phase separation mechanisms[6], according to different and distinct combinations of T and ΦPCL at different locations of the combinatorial library chip, PLGA and PCL phases were separated to form sur
Data Loading...
