The gut microbiota is composed of a large number of different bacteria, that play a key role in the construction of a metabolic signaling network. Deepening the link between metabolic pathways of the gut microbiota and human health, it seems increasingly essential to evolutionarily define the principal technologies applied in the field and their future trends. We use a topic analysis tool, Latent Dirichlet Allocation, to extract themes as a probabilistic distribution of latent topics from literature dataset. We also use the Prophet neural network prediction tool to predict future trend of this area of study. A total of 1,271 abstracts (from 2006 to 2020) were retrieved from MEDLINE with the query on "gut microbiota" and "metabolic pathway." Our study found 10 topics covering current research types: dietary health, inflammation and liver cancer, fatty and diabetes, microbiota community, hepatic metabolism, metabolomics-based approach and SFCAs, allergic and immune disorders, gut dysbiosis, obesity, brain reaction, and cardiovascular disease. The analysis indicates that, with the rapid development of gut microbiota research, the metabolomics-based approach and SCFAs (topic 6) and dietary health (topic 1) have more studies being reported in the last 15 years. We also conclude from the data that, three other topics could be heavily focused in the future: metabolomics-based approach and SCFAs (topic 6), obesity (topic 8) and brain reaction and cardiovascular disease (topic 10), to unravel microbial affecting human health.
Keywords: LDA; gut microbiota; metabolic pathway; time-series feature; topic prediction.
Copyright © 2020 Ning, Lifang and Huixin.