Fig. 5

Recipient effect on post-FMT predictions. A SCCs of different orders and Shannon reported over the ABX-treated mice (purple bars) vs the GF mice (light blue bars) compared to the overall raw donor-recipient correlations (gray).B SCCs of different orders and Shannon reported over the ABX treated cohorts (light bars) vs the no-ABX, untreated cohorts (dark bars) compared to the overall raw donor-recipient correlations (gray). C, D Similarity Euclidean distance between two recipients preprocessed ASVs (order-level) vectors in the human-to-mouse cohorts (GF and ABX) (D) and the human-to-human cohorts (ABX vs no ABX). In each pair, the rightmost bar represents the distances between samples of different recipients that got FMT from the same donor , SDDR (measures the effect of the recipient background), and the leftmost bar represents the distances between samples of different recipients that received FMT from different donors, DDDR (measures the effect of the donor), with a significant hierarchy of distances. The lowest distances are within the same donor/recipient, followed by the same donor, followed by different donors (\(*\) \(p<0.05\),\(**\) \(p<0.01\), \(***\) \(p<0.001\)). E, F Comparison of FMT effect in the mouse cohorts in GF mice (E) and in ABX mice (F) at different times. In the GF mice, there is a donor effect at 7D and not at 28D. Similarly, in the ABX-treated cohort, there is a difference at 10W (1W post-FMT) and not at 15W (6 weeks post-FMT). G Comparison of SCCs between the recipient’s properties post-FMT with the donor (orange), the recipient before the FMT (gray), and the predicted recipient property post-FMT by iMic (pink) in the human-to-human cohorts. The prediction is typically much higher than the two others