Background: The relationship among the human gut microbiome, microbially produced metabolites, and health outcomes remains of great interest. To decrease participant burden, room-temperature storage methods for fecal samples have become increasingly important. However, kits for storing the fecal microbiome and metabolome have not been well explored. We hypothesized that storing fecal samples by drying them with silica gel may be suitable. Objectives: The objective was to evaluate the performance of storage at room temperature by drying feces for subsequent examination of the microbiome, microbial pathways, and the metabolome. Methods: Feces from ten healthy adults (6 male and 4 female) were sampled and immediately processed, as controls, and stored at room temperature in an incubator, on an FTA card, in RNAlater, or dried by silica gel. Storage at room temperature continued for 7 days. Drying by the silica gel method was assessed for 14 days. The fecal microbiome was assessed by sequencing the bacterial 16S ribosomal RNA-encoding gene (V1-V2 region), fecal microbial pathway profiles were analyzed by whole-genome shotgun metagenomics, and fecal metabolome profiles were analyzed using capillary electrophoresis time-of-flight mass spectrometry (CE-TOFMS). Results: Qualitative and β-diversity analyses of the microbiome, microbial pathways, and the metabolome showed that drying by silica gel were closest to those immediately after processing. When focusing on the abundances of individual microbes, microbial pathways, and metabolites, some were found to be significantly different. However, the intra-method ranking of individual items showed that 100%, 87-97%, and 63-76% of microbes, microbial pathways, and metabolites, respectively, were significantly correlated between silica gel preserving and immediately processing method.

Recent advances in microbiome research have led to the further development of microbial interventions, such as probiotics and prebiotics, which are potential treatments for constipation. However, the effects of probiotics vary from person to person; therefore, the effectiveness of probiotics needs to be verified for each individual. Individuals showing significant effects of the target probiotic is called responders. A statistical model for the evaluation of responders was proposed in a previous study. However, the previous model does not consider the lag in the effect of the probiotic. It is expected that there are lags between the period of time when probiotics are administered and when they are effective. In this study, we propose a Bayesian statistical model to estimate the probability that a subject is a responder, by considering the lag of the effect period. In synthetic dataset experiments, the proposed model was found to outperform the base model, which did not factor in the lag. Further, we found that the proposed model could distinguish responders showing large uncertainty in terms of the lag of the effect period against the intake period.