The human gut microbiome and its influence in mental health
Mental health problems affect 25% of the population, making it the leading cause of disability globally. Normal microbiota is related with healthy states, however changes in its composition (called dysbiosis) is linked with non-healthy pathologies. In this project we explored the influence of the gu...
| Autor: | |
|---|---|
| Tipo de recurso: | tesis de maestría |
| Fecha de publicación: | 2020 |
| País: | España |
| Institución: | Universitat Oberta de Catalunya (UOC) |
| Repositorio: | O2, repositorio institucional de la UOC |
| OAI Identifier: | oai:openaccess.uoc.edu:10609/106367 |
| Acceso en línea: | http://hdl.handle.net/10609/106367 |
| Access Level: | acceso abierto |
| Palabra clave: | gut microbiome machine learning gut-brain axis mental health microbiota intestinal aprendizaje automático salud mental eje cerebro-intestino aprenentatge automàtic salut mental eix cervell-intestí Bioinformatics -- TFM Bioinformàtica -- TFM Bioinformática -- TFM |
| Sumario: | Mental health problems affect 25% of the population, making it the leading cause of disability globally. Normal microbiota is related with healthy states, however changes in its composition (called dysbiosis) is linked with non-healthy pathologies. In this project we explored the influence of the gut microbiota on the occurrence of non-healthy mental states using machine learning approaches, enterotype classifications and univariate and multivariate statistical analyses. Among the demographic characteristics we found differences between the mental illness states in the ethnicity (p = 0.04), sex (p = 0.004), irritable bowel disease (p < 0.001), etc. We found lower levels of Firmicutes and higher levels of Bacteroidetes and a lower Firmicutes/Bacteroidetes ratio in the gut of people with mental issues. People with mental illness has a lower alpha diversity index in their gut in comparison with healthy people (p = 0.002). The beta diversity analysis presented different centroids regarding the mental states statistically measured by the PERMANOVA test (p = 0.032). The best machine learning predictor was Random Forest with an accuracy of 0.62. However, because of we mixed the different mental disorders with a different biological background, probably creating noise, the prediction results of the machine learning algorithms do not have better performances. In conclusion, more efforts are necessary in the use of machine learning algorithms with microbiome information, because of the potential that these methods have in the classification and/or prediction of certain pathologies. Also, higher Firmicutes and lower Bacteroidetes could be risk factor in the occurrence of a mental illness. |
|---|