G2S: A New Deep Learning Tool for Predicting Stool Microbiome Structure From Oral Microbiome Data

Deep learning methodologies have revolutionized prediction in many fields and show the potential to do the same in microbial metagenomics. However, deep learning is still unexplored in the field of microbiology, with only a few software designed to work with microbiome data. Within the meta-community theory, we foresee new perspectives for the development and application of deep learning algorithms in the field of the human microbiome. In this context, we developed G2S, a bioinformatic tool for taxonomic prediction of the human fecal microbiome directly from the oral microbiome data of the same individual. The tool uses a deep convolutional neural network trained on paired oral and fecal samples from populations across the globe, which allows inferring the stool microbiome at the family level more accurately than other available approaches. The tool can be used in retrospective studies, where fecal sampling was not performed, and especially in the field of paleomicrobiology, as a unique opportunity to recover data related to ancient gut microbiome configurations. G2S was validated on already characterized oral and fecal sample pairs, and then applied to ancient microbiome data from dental calculi, to derive putative intestinal components in medieval subjects.

Marine Bacteria Community in a 150-m Depth Tachai Island, the Southeast Andaman Sea of Thailand

The southeast Andaman Sea 52-m off the west coast of Phang Nga province, Thailand, is located in the Indian Ocean, representing a hotspot for marine biodiversity of the world. This study utilized metagenomics combined 16S rRNA gene (V3–V4) sequencing, and firstly revealed the microbiota and their metabolism potentials of this site at an epipelagic depth (150-m depth, TC150M), including comparison with its pelagic depth (30-m depth, TC30M) as well as other nearby Thailand and global ocean sites. Between TC150M and TC30M, the TC150M microbial metagenomics was an eight-fold higher, and the microbiota comprised, for examples more abundant Bacteroidetes while fewer Proteobacteria, than the TC30M. The microbial metabolic potentials of the TC150M were statistically higher in replication repair and metabolisms of amino acids, lipids, nucleotides, and xenobiotics biodegradation, etc. Following comparative microbiota analyses between three Andaman Sea sites and two Gulf of Thailand sites, the relatively great proportions of Bacteroidetes, Nitrospirae, Gemmatimonadetes, and Chlorobi characterized the southeast Andaman Sea. Nevertheless, the microbiota representing Thailand marine sites remained distinguished from the global ocean sites where beta diversities were close. Thai maritime sites showed proportionally higher Proteobacteria, Bacteroides, Nitrospirae, Gemmatimonadetes, and Chlorobi. Thus, the Thai marine microbiota database helps better understand our global ocean microbiota and microbial metabolic potentials. Here, the microbial metabolism potentials between Thailand and the global ocean sites of relatively close microbiota databases encompose the similar functions yet in statistically different frequencies. Our research provided the first preliminary marine microbiome comparison between the epipelagic and pelagic sea levels of the southeast Andaman Sea, Thailand.

G2S: a new deep learning tool for predicting stool microbiome structure from oral microbiome data

Background Deep learning methodologies have revolutionized prediction in many fields and show the potential to do the same in microbial metagenomics. However, deep learning is still unexplored in the field of microbiology, with only a few software designed to work with microbiome data. In the frame of meta-community theory, we foresee new perspectives for the development and application of deep learning algorithms in microbiology, with a great potential in the field of human microbiome. Results G2S is a bioinformatic tool for the taxonomic prediction of the human stool microbiome directly from oral microbiome data of the same individual. The tool uses a deep convolutional neural network trained on data of the Human Microbiome Project, allowing to infer the stool microbiome at the family level more accurately than other approaches. G2S was validated on already characterized oral and fecal sample pairs, and then applied to ancient microbiome data from dental calculi, to derive putative intestinal components in medieval subjects. Conclusions G2S infers the family-level taxonomic configuration of the stool microbiome mirroring the real composition with exceptional performance. G2S can be used with modern samples, allowing to predict the eubiotic/dysbiotic state of the gut microbiome when fecal sampling is missing, and especially with ancient samples, as a unique opportunity in the field of paleomicrobiology to recover data related to ancient gut microbiome configurations.