Age signature in the bacterial structure: Researchers use the characteristics of the microbial communities in and on a person’s body to estimate their age. Your machine learning method can determine the age of skin smears with an accuracy of about 3.8 years, followed by 4.5 years for mouth samples and 11.5 years for stool samples. According to the scientists, their process has the potential to research aging processes and diseases related to microbial colonization patterns.
It is teeming with us and with us: Man is the habitat of a motley community of many different types of microbes. In recent years, the importance of these residents has come into focus. Numerous studies have already shown how important a cheap microbiome is for our health. In the intestine, mouth and on the skin, for example, friendly types of bacteria prevent pathogens from spreading. In addition, the tiny things also have an impact on our immune system and other processes in the body. For example, certain compositions of the microbial community have been linked to autoimmune diseases, obesity, and even mental health problems. However, the importance of a healthy microbiome still has a lot of research potential, experts say.
Typical age patterns are emerging
It is already known that numerous factors influence the composition of the communities in the intestine, mouth or on the skin. In addition to nutrition, different habits, living conditions and our disposition, age also influences the microbiomes. A team of US scientists has now dealt with the last aspect in more detail. The study was initiated in 2014 by a study conducted by researchers who examined the microbiome of infant stool samples in the first few months of life. They found that the maturity of the children was reflected in the bacterial compositions. In the current study, the scientists now wanted to find out to what extent these links can also be transferred to adults of different ages and to the microbes of other parts of the body.
They used microbiome sequencing data from public databases and from several citizen science projects from different countries as a source of information, in which microbiome measurements of stool, saliva or skin smears from the participants were recorded. The study is based on 4434 stool samples, 2550 saliva samples and 1975 skin samples. The participants, whose data were used in the study, were between 18 and 90 years old and had no illnesses or other circumstances that could distort the result. As the researchers explain, they fed the data to a special computer program that is capable of machine learning: through training, the intelligent system can recognize certain patterns in data in order to link them to other variables – in this case, age.
Procedures with medical potential
As the researchers report, their system was finally able to use the microbiome samples to estimate the chronological age of a person with varying degrees of accuracy. Skin samples gave the most accurate prediction with plus-minus 3.8 years, followed by the oral samples with 4.5 years and 11.5 years with the stool samples. The recognition criterion was the diversity and frequency of the types of microbes, which tends to disappear with increasing age, the scientists explain. One possible reason why the microbes living on our skin change so significantly is, according to the researchers, the change in skin physiology over the years: For certain ages, specific characteristics of talc production and moisture are typical, which affect microbial colonization ,
As the researchers emphasize, their study is now initially a fundamental demonstration of the potential of the process. “Ways to correlate microbes with age could help us uncover the role of microbes in the aging process and develop treatments for age-related diseases that target microbiomes,” said co-author Zhenjiang Zech Xu of the University of California at San Diego , His colleague Ho-Cheol Kim adds: “Our results show the potential of using machine learning and artificial intelligence to better understand human microbiomes. The application of this technology in future studies could help to gain deeper insights into the connection of microbial societies with our health and many diseases, ”said the scientist.
Source: University of California – San Diego, technical article: mSystems, doi: 10.1128 / mSystems.00630-19