Dataset Information

Reduction of the Number of Samples for Cost-Effective Hyperspectral Grape Quality Predictive Models.

ABSTRACT: Developing chemometric models from near-infrared (NIR) spectra requires the use of a representative calibration set of the entire population. Therefore, generally, the calibration procedure requires a large number of resources. For that reason, there is a great interest in identifying the most spectrally representative samples within a large population set. In this study, principal component and hierarchical clustering analyses have been compared for their ability to provide different representative calibration sets. The calibration sets generated have been used to control the technological maturity of grapes and total phenolic compounds of grape skins in red and white cultivars. Finally, the accuracy and precision of the models obtained with these calibration sets resulted from the application of the selection algorithms studied have been compared with each other and with the whole set of samples using an external validation set. Most of the standard errors of prediction (SEP) in external validation obtained from the reduced data sets were not significantly different from those obtained using the whole data set. Moreover, sample subsets resulting from hierarchical clustering analysis appear to produce slightly better results.

SUBMITTER: Nogales-Bueno J

PROVIDER: S-EPMC7912666 | biostudies-literature | 2021 Jan

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Reduction of the Number of Samples for Cost-Effective Hyperspectral Grape Quality Predictive Models.

Nogales-Bueno Julio J Rodríguez-Pulido Francisco José FJ Baca-Bocanegra Berta B Pérez-Marin Dolores D Heredia Francisco José FJ Garrido-Varo Ana A Hernández-Hierro José Miguel JM

Foods (Basel, Switzerland) 20210123 2

Developing chemometric models from near-infrared (NIR) spectra requires the use of a representative calibration set of the entire population. Therefore, generally, the calibration procedure requires a large number of resources. For that reason, there is a great interest in identifying the most spectrally representative samples within a large population set. In this study, principal component and hierarchical clustering analyses have been compared for their ability to provide different representa ...[more]

PMID: 33498776

Similar Datasets

Project description:BackgroundDespite the extensive use of bevacizumab in a range of oncology indications, the US FDA revoked its approval for breast cancers, and multiple negative trials in several solid malignancies have been reported, so the need for predictive biomarkers has increased. The development of predictive biomarkers for anti-angiogenic bevacizumab therapy has long been pursued but without success.IntroductionHeat shock protein (HSP)-27 expression has recently been identified as a predictive biomarker for bevacizumab in treating metastatic melanoma. This study aimed to evaluate the cost effectiveness of HSP27 biomarker testing before administration of bevacizumab.MethodsA partitioned survival analysis model with three mutually exclusive health states (progression-free survival, progressed disease, and death) was developed using a Norwegian health system perspective. The proportion of patients in each state was calculated using the area under the Kaplan-Meier curve for progression-free and overall survival derived from trials of bevacizumab and dacarbazine. Three strategies were compared: (1) test-treat with HSP27 biomarker and bevacizumab, (2) treat-all with dacarbazine without HSP27 testing, (3) treat-all with bevacizumab without HSP27 testing. Quality-adjusted life-years (QALYs) and costs were calculated for each strategy and discounted at 4%. A lifetime horizon was applied. Uncertainty analyses were performed. Expected value of perfect information (EVPI) was estimated to assess the potential value of further research to generate more evidence.ResultsAlthough the test-treat strategy was cost effective compared with treat-all with dacarbazine, it was not cost effective compared with treat-all with bevacizumab without HSP27 testing. However, EVPI results showed very minimal or no value in conducting further research efforts to reduce uncertainties around current information.ConclusionThe results of this study suggested that testing for HSP27 expression before administering bevacizumab is not cost effective compared with treat-all with bevacizumab without testing. It indicates that HSP27 expression is not cost effective as a potential predictive biomarker for bevacizumab. This may not necessarily mean that HSP27 is a bad biomarker for bevacizumab, but it may mean that bevacizumab is much better than dacarbazine regardless of HSP27 expression, so patient stratification according to HSP27 status is meaningless. Or, indeed, it may imply that HSP27 is not sufficiently good at identifying the right patients for bevacizumab.

Dataset Information

Reduction of the Number of Samples for Cost-Effective Hyperspectral Grape Quality Predictive Models.

Publications

Reduction of the Number of Samples for Cost-Effective Hyperspectral Grape Quality Predictive Models.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets