Researchers draw connections between microbiome and human traits from study involving groups from Israel and the U.S.

Researchers at the Weizmann Institute of Science explored the connections between elements of the microbiome and biomarkers of human traits through a cohort study including populations from Israel and the United States.

Associations between human conditions (diabetes, obesity and other metabolic disorders) and the microbiome have been inconsistent likely due to insufficient sample sizes. In this study, researchers collected samples from 34,057 individuals from both countries and found robust associations between the microbiome and human traits.    

"We show that a large fraction of the variance of several traits such as age, HbA1C% and BMI can be accurately predicted by both linear models and boosting decision trees models," researchers wrote in a preprint of “An atlas of robust microbiome associations with phenotypic traits based on large-scale cohorts from two continents” published at BioRxiv on May 30, 2020.    

They demonstrate that subsampling even as many as 1000 individuals results in much less accuracy for associations between the microbiome and human disease.

Interestingly some of the bacteria they identified as top hits for associations with human traits were unnamed in many databases. They say the results may explain why few different studies on bacteria and human traits are in agreement. This caused them to state the importance of the continued effort to increase the reference set for microbial genomes.

Researchers analyzed participant provided stool samples and used PowerMag Soil DNA isolation kit to purify the Genomic DNA. Samples were sequenced on an Illumina Nextera 500 or NovaSeq 6000. Trimmomatic was used to ensure reads with Illumina adapters were not included and to filter out low quality reads.