Wang Xun, Sun Beibei, Liu Boyang, Fu Yaping, Zheng Pan
College of Computer and Communication Engineering, China University of Petroleum, Qingdao 266580, Shandong, China.
State-owned Asset and Laboratory Management Department, China University of Petroleum, Qingdao 266580, Shandong, China.
PLoS One. 2017 Nov 2;12(11):e0186853. doi: 10.1371/journal.pone.0186853. eCollection 2017.
Experimental design focuses on describing or explaining the multifactorial interactions that are hypothesized to reflect the variation. The design introduces conditions that may directly affect the variation, where particular conditions are purposely selected for observation. Combinatorial design theory deals with the existence, construction and properties of systems of finite sets whose arrangements satisfy generalized concepts of balance and/or symmetry. In this work, borrowing the concept of "balance" in combinatorial design theory, a novel method for multifactorial bio-chemical experiments design is proposed, where balanced templates in combinational design are used to select the conditions for observation. Balanced experimental data that covers all the influencing factors of experiments can be obtianed for further processing, such as training set for machine learning models. Finally, a software based on the proposed method is developed for designing experiments with covering influencing factors a certain number of times.
实验设计侧重于描述或解释那些被假设为反映变异的多因素相互作用。该设计引入可能直接影响变异的条件，其中特意选择特定条件进行观察。组合设计理论处理有限集系统的存在性、构造和性质，这些有限集的排列满足平衡和/或对称的广义概念。在这项工作中，借鉴组合设计理论中的“平衡”概念，提出了一种用于多因素生化实验设计的新方法，其中组合设计中的平衡模板用于选择观察条件。可以获得涵盖实验所有影响因素的平衡实验数据，以便进行进一步处理，例如用于机器学习模型的训练集。最后，基于所提出的方法开发了一个软件，用于设计能使影响因素被覆盖一定次数的实验。
