Suyang Wu, Stephanie A. Ketcham, Claudia Corredor, Douglas Both, Yuxiang Zhao, James K. Drennen, Carl A. Anderson
{"title":"通过数据融合策略优化自适应建模:利用电容光谱实时预测凋亡细胞百分比。","authors":"Suyang Wu, Stephanie A. Ketcham, Claudia Corredor, Douglas Both, Yuxiang Zhao, James K. Drennen, Carl A. Anderson","doi":"10.1002/btpr.3424","DOIUrl":null,"url":null,"abstract":"<p>The previous research showcased a partial least squares (PLS) regression model accurately predicting cell death percentages using in-line capacitance spectra. The current study advances the model accuracy through adaptive modeling employing a data fusion approach. This strategy enhances prediction performance by incorporating variables from the Cole–Cole model, conductivity and its derivatives over time, and Mahalanobis distance into the predictor matrix (X-matrix). Firstly, the Cole–Cole model, a mechanistic model with parameters linked to early cell death onset, was integrated to enhance prediction performance. Secondly, the inclusion of conductivity and its derivatives over time in the X-matrix mitigated prediction fluctuations resulting from abrupt conductivity changes during process operations. Thirdly, Mahalanobis distance, depicting spectral changes relative to a reference spectrum from a previous time point, improved model adaptability to independent test sets, thereby enhancing performance. The final data fusion model substantially decreased root-mean squared error of prediction (RMSEP) by around 50%, which is a significant boost in prediction accuracy compared to the prior PLS model. Robustness against reference spectrum selection was confirmed by consistent performance across various time points. In conclusion, this study illustrates that the data fusion strategy substantially enhances the model accuracy compared to the previous model relying solely on capacitance spectra.</p>","PeriodicalId":8856,"journal":{"name":"Biotechnology Progress","volume":null,"pages":null},"PeriodicalIF":2.5000,"publicationDate":"2024-01-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Adaptive modeling optimized by the data fusion strategy: Real-time dying cell percentage prediction using capacitance spectroscopy\",\"authors\":\"Suyang Wu, Stephanie A. Ketcham, Claudia Corredor, Douglas Both, Yuxiang Zhao, James K. Drennen, Carl A. Anderson\",\"doi\":\"10.1002/btpr.3424\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"<p>The previous research showcased a partial least squares (PLS) regression model accurately predicting cell death percentages using in-line capacitance spectra. The current study advances the model accuracy through adaptive modeling employing a data fusion approach. This strategy enhances prediction performance by incorporating variables from the Cole–Cole model, conductivity and its derivatives over time, and Mahalanobis distance into the predictor matrix (X-matrix). Firstly, the Cole–Cole model, a mechanistic model with parameters linked to early cell death onset, was integrated to enhance prediction performance. Secondly, the inclusion of conductivity and its derivatives over time in the X-matrix mitigated prediction fluctuations resulting from abrupt conductivity changes during process operations. Thirdly, Mahalanobis distance, depicting spectral changes relative to a reference spectrum from a previous time point, improved model adaptability to independent test sets, thereby enhancing performance. The final data fusion model substantially decreased root-mean squared error of prediction (RMSEP) by around 50%, which is a significant boost in prediction accuracy compared to the prior PLS model. Robustness against reference spectrum selection was confirmed by consistent performance across various time points. In conclusion, this study illustrates that the data fusion strategy substantially enhances the model accuracy compared to the previous model relying solely on capacitance spectra.</p>\",\"PeriodicalId\":8856,\"journal\":{\"name\":\"Biotechnology Progress\",\"volume\":null,\"pages\":null},\"PeriodicalIF\":2.5000,\"publicationDate\":\"2024-01-04\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Biotechnology Progress\",\"FirstCategoryId\":\"5\",\"ListUrlMain\":\"https://onlinelibrary.wiley.com/doi/10.1002/btpr.3424\",\"RegionNum\":3,\"RegionCategory\":\"生物学\",\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q3\",\"JCRName\":\"BIOTECHNOLOGY & APPLIED MICROBIOLOGY\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Biotechnology Progress","FirstCategoryId":"5","ListUrlMain":"https://onlinelibrary.wiley.com/doi/10.1002/btpr.3424","RegionNum":3,"RegionCategory":"生物学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q3","JCRName":"BIOTECHNOLOGY & APPLIED MICROBIOLOGY","Score":null,"Total":0}
引用次数: 0
摘要
之前的研究展示了一个偏最小二乘(PLS)回归模型,该模型能利用在线电容谱准确预测细胞死亡百分比。当前的研究通过采用数据融合方法的自适应建模,提高了模型的准确性。该策略将科尔-科尔模型、电导率及其随时间变化的导数以及马哈拉诺比距离等变量纳入预测矩阵(X 矩阵),从而提高了预测性能。首先,整合了科尔-科尔模型,这是一个包含与细胞早期死亡相关参数的机理模型,以提高预测性能。其次,在 X 矩阵中加入电导率及其随时间变化的导数,可减轻因工艺操作过程中电导率突然变化而导致的预测波动。第三,Mahalanobis 距离描述了相对于前一个时间点的参考光谱的光谱变化,提高了模型对独立测试集的适应性,从而提高了性能。最终的数据融合模型将预测的均方根误差(RMSEP)大幅降低了约 50%,与之前的 PLS 模型相比,预测准确度有了显著提高。不同时间点上的一致表现证实了参考谱选择的鲁棒性。总之,这项研究表明,与之前仅依赖电容谱的模型相比,数据融合策略大大提高了模型的准确性。
Adaptive modeling optimized by the data fusion strategy: Real-time dying cell percentage prediction using capacitance spectroscopy
The previous research showcased a partial least squares (PLS) regression model accurately predicting cell death percentages using in-line capacitance spectra. The current study advances the model accuracy through adaptive modeling employing a data fusion approach. This strategy enhances prediction performance by incorporating variables from the Cole–Cole model, conductivity and its derivatives over time, and Mahalanobis distance into the predictor matrix (X-matrix). Firstly, the Cole–Cole model, a mechanistic model with parameters linked to early cell death onset, was integrated to enhance prediction performance. Secondly, the inclusion of conductivity and its derivatives over time in the X-matrix mitigated prediction fluctuations resulting from abrupt conductivity changes during process operations. Thirdly, Mahalanobis distance, depicting spectral changes relative to a reference spectrum from a previous time point, improved model adaptability to independent test sets, thereby enhancing performance. The final data fusion model substantially decreased root-mean squared error of prediction (RMSEP) by around 50%, which is a significant boost in prediction accuracy compared to the prior PLS model. Robustness against reference spectrum selection was confirmed by consistent performance across various time points. In conclusion, this study illustrates that the data fusion strategy substantially enhances the model accuracy compared to the previous model relying solely on capacitance spectra.
期刊介绍:
Biotechnology Progress , an official, bimonthly publication of the American Institute of Chemical Engineers and its technological community, the Society for Biological Engineering, features peer-reviewed research articles, reviews, and descriptions of emerging techniques for the development and design of new processes, products, and devices for the biotechnology, biopharmaceutical and bioprocess industries.
Widespread interest includes application of biological and engineering principles in fields such as applied cellular physiology and metabolic engineering, biocatalysis and bioreactor design, bioseparations and downstream processing, cell culture and tissue engineering, biosensors and process control, bioinformatics and systems biology, biomaterials and artificial organs, stem cell biology and genetics, and plant biology and food science. Manuscripts concerning the design of related processes, products, or devices are also encouraged. Four types of manuscripts are printed in the Journal: Research Papers, Topical or Review Papers, Letters to the Editor, and R & D Notes.