Metagenomic Thermometer

AbstractVarious microorganisms exist in environments, and each of which has an optimal growth temperature (OGT). The relationship between genomic information and OGT of each species has long been studied, and one such study revealed that OGT of prokaryotes can be accurately predicted based on the fraction of seven amino acids (IVYWREL) among all encoded amino-acid sequences in its genome. Extending this discovery, the fraction of IVYWREL among all amino acids encoded in the metagenomic DNA in an environment may reflect the environmental temperature to which the microorganisms have adapted. In the present study, we developed a ‘Metagenomic Thermometer’ as a means of predicting environmental temperature based on metagenomic sequences. Analysis of 41 metagenomic samples obtained from hot springs of varying temperature revealed that the Metagenomic Thermometer could indeed predict environmental temperature with high accuracy (R2 = 0.75). Although the taxonomic composition of microbial community was substantially affected by pH, the performance of the Metagenomic Thermometer was refractory to changes in community composition. The Metagenomic Thermometer was also applied to 616 human gut metagenomic samples, yielding a reasonably accurate value (i.e., 36.22 ± 1.98°C) for human body temperature. Our results suggest that environmental temperature can be predicted based on metagenomic sequence data.