Scholarlink

<p>Low atmospheric pressure is a major environmental factor in space that can affect the physiological responses of plants. Artificial neural networks (ANNs), along with optimization algorithms, can be beneficial computational approaches to model and optimize seedling growth under in vitro conditions. This study aimed to model and optimize morphological responses of in vitro chamomile (<i>Matricaria chamomilla</i> L.) seedlings to low atmospheric pressures, population, and developmental stage by ANNs and genetic algorithms for the first time. The Multilayer Perceptron (MLP) was identified as the best model for predicting morphological parameters and flavonoid content based on the inclusion of specific comparative performance using R2, RMSE, MAE, and MBE. The sensitivity analysis showed an important impact of low pressure on the morphological responses and flavonoid content, followed by population. MLP linked to Non-dominated Sorting Genetic Algorithm-II (NSGA-II) optimized morphological traits and flavonoid content and predicted that the sprout priming with 350 mbar pressure in the Ardabil population would result in the highest values for FW (0.46&#xa0;g), DW (0.043&#xa0;g), RWC (95%), aerial part height (6.59&#xa0;cm), root length (9.78&#xa0;cm), root number (9.4), and flavonoid (639.8&#xa0;µg g<sup>− 1</sup>) content. For validation, new sprouts of the Ardabil population were cultivated under 350 mbar, and a considerable enhancement in FW (104.5%), DW (95.2%), aerial part height (200.4%), root length (145.2%), root number (312.5%), and flavonoid (63.6%) were observed comparing to the control. Findings showed that MLP-NSGAII acts as a powerful framework to establish new computational strategies for plant in vitro cultivation under stress conditions.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Modeling and optimizing chamomile morphology and flavonoids by Multilayer Perceptron coupled with Non-dominated Sorting Genetic Algorithm-II under low atmospheric pressure

  • 期刊论文

  • 发表日期：

    2026年6月1日

查看全文
  • Halimeh Hassanpour,
  • Mahsa Sardari,
  • Shima Tabibian

摘要

Low atmospheric pressure is a major environmental factor in space that can affect the physiological responses of plants. Artificial neural networks (ANNs), along with optimization algorithms, can be beneficial computational approaches to model and optimize seedling growth under in vitro conditions. This study aimed to model and optimize morphological responses of in vitro chamomile (Matricaria chamomilla L.) seedlings to low atmospheric pressures, population, and developmental stage by ANNs and genetic algorithms for the first time. The Multilayer Perceptron (MLP) was identified as the best model for predicting morphological parameters and flavonoid content based on the inclusion of specific comparative performance using R2, RMSE, MAE, and MBE. The sensitivity analysis showed an important impact of low pressure on the morphological responses and flavonoid content, followed by population. MLP linked to Non-dominated Sorting Genetic Algorithm-II (NSGA-II) optimized morphological traits and flavonoid content and predicted that the sprout priming with 350 mbar pressure in the Ardabil population would result in the highest values for FW (0.46 g), DW (0.043 g), RWC (95%), aerial part height (6.59 cm), root length (9.78 cm), root number (9.4), and flavonoid (639.8 µg g− 1) content. For validation, new sprouts of the Ardabil population were cultivated under 350 mbar, and a considerable enhancement in FW (104.5%), DW (95.2%), aerial part height (200.4%), root length (145.2%), root number (312.5%), and flavonoid (63.6%) were observed comparing to the control. Findings showed that MLP-NSGAII acts as a powerful framework to establish new computational strategies for plant in vitro cultivation under stress conditions.