Fecal transplantation, or more precisely fecal microbiota transplantation (FMT), can cure Clostridium difficile (“Cdiff”) infections of the GI tract. Unfortunately, the efficacy of this treatment is quite variable. Eric Alm’s group now developed a new method, based on whole-genome shotgun sequencing (WGS) of stool samples, to track bacterial strains from the donor in the recipient. The results of this exercise were used to train a machine learning based model in order to determine the factors that determine success of donor bacterial strains when transplanted into hosts. Perhaps not too surprisingly, absolute abundance of a strain in the donor sample was the most significant predictor of success. Also perhaps not too surprisingly, the next important factor was the group of bacteria different strains belonged to. Why some bacteria are more successful in FMT, and whether improving the success of certain bacteria may improve the efficacy of FTM treatment are important questions for future research.
comment written by Detlef Weigel