1. Exploring the Impact of Food on the Gut Ecosystem Based on the Combination of Machine Learning and Network Visualization
- Author
-
Yoshihiro Inoue, Taisei Kanamoto, Hideaki Shima, Amiu Shino, Mizuho Kajikawa, Jun Kikuchi, Yasuhiro Date, Shizuka Masuda, and Yuuri Tsuboi
- Subjects
0301 basic medicine ,Male ,food intake ,Matteuccia ,gut ecosystem ,lcsh:TX341-641 ,Machine learning ,computer.software_genre ,metabolic response ,Models, Biological ,Article ,law.invention ,Functional networks ,03 medical and health sciences ,Probiotic ,0302 clinical medicine ,law ,Evaluation methods ,Humans ,Ecosystem ,network analysis ,Nutrition and Dietetics ,biology ,business.industry ,Microbiota ,Probiotics ,biology.organism_classification ,Highly sensitive ,Gastrointestinal Tract ,030104 developmental biology ,Prebiotics ,machine learning ,Food ,Ferns ,Pteridium aquilinum ,Artificial intelligence ,business ,computer ,Host (network) ,lcsh:Nutrition. Foods and food supply ,030217 neurology & neurosurgery ,Food Science - Abstract
Prebiotics and probiotics strongly impact the gut ecosystem by changing the composition and/or metabolism of the microbiota to improve the health of the host. However, the composition of the microbiota constantly changes due to the intake of daily diet. This shift in the microbiota composition has a considerable impact; however, non-pre/probiotic foods that have a low impact are ignored because of the lack of a highly sensitive evaluation method. We performed comprehensive acquisition of data using existing measurements (nuclear magnetic resonance, next-generation DNA sequencing, and inductively coupled plasma-optical emission spectroscopy) and analyses based on a combination of machine learning and network visualization, which extracted important factors by the Random Forest approach, and applied these factors to a network module. We used two pteridophytes, Pteridium aquilinum and Matteuccia struthiopteris, for the representative daily diet. This novel analytical method could detect the impact of a small but significant shift associated with Matteuccia struthiopteris but not Pteridium aquilinum intake, using the functional network module. In this study, we proposed a novel method that is useful to explore a new valuable food to improve the health of the host as pre/probiotics.
- Published
- 2017