Only 4.7% of apparent FMT colonization events involve actual donor strain replacement (Aug 2026, n=13) Strain-level ecological filtering governs microbial colonization of the human gut FMT

[IsthmusSIBO]

https://biomesci.com/fmt-strain-level-colonization-cell-reports/
https://www.cell.com/cell-reports/fulltext/S2211-1247(26)00853-3

Researchers tracked 13 transplant patients and their donors using genome-level sequencing, going deeper than the usual species-level test. They reconstructed thousands of bacterial genomes and compared them at single-nucleotide resolution, precise enough to tell whether a bacterium in the patient's gut is genuinely the donor's exact strain (a specific genetic lineage) or just a similar strain the patient already had.

Turns out a lot of "successful colonization" was an illusion. Donor bacterial species often showed up on standard tests, but the patient's own resident strain of that same species had usually just rebounded, not the donor's strain taking over.

Click to shrink...

Summary​

Microbial colonization of the human gut is typically inferred from species-level profiling, yet durable establishment operates at the strain level. Here, using longitudinal shotgun metagenomics across multiple donor-recipient pairs undergoing fecal microbiota transplantation, we show that colonization is governed by lineage-dependent strain-level ecological filtering. Strain-resolved analyses reveal that gut colonization imposes reproducible population-genetic bottlenecks, characterized by reduced nucleotide diversity and selective strain capture. Lineage identity is the primary determinant of strain fate: certain taxa exhibit high donor-strain fidelity, whereas dominant gut lineages, most notably Lachnospiraceae, display broad species-level engraftment but limited capture of donor-identical strains. Repeated transplantation progressively increases species-level retention, building ecological memory, yet fails to overcome intrinsic barriers to consensus-level donor-strain capture. Clinical remission aligned specifically with directional donor-strain replacement rather than taxonomic remodeling alone, identifying strain-level lineage compatibility as a candidate determinant of therapeutic success. Collectively, these findings establish that gut colonization is constrained by strain-level ecological filtering and reframe microbiota transplantation as a selective evolutionary process in which lineage identity, not inoculum diversity, gates therapeutic integration.

Click to shrink...

Key takeaways

• Researchers tracked 13 fecal transplant patients and their donors with genome-level (strain-resolved) sequencing, not just species-level testing.
• Donor bacterial species often showed up in the recipient's gut, but the exact donor strain (the specific genetic lineage within that species) rarely took hold. Only 22 of 465 apparent colonization events were confirmed as true donor-strain capture.
• Clinical remission tracked with donor-strain replacement specifically, not with how much the overall gut community resembled the donor's.
• This is an observational finding in FMT patients, not a randomized trial, so it points to a mechanism worth testing, not a guaranteed fix.

Click to shrink...

 

[SFBayFMT5]

This is also interesting for the following statement, which begins the second paragraph after the subheading "Repeated FMT exposes ecological hysteresis and a minimal persistent donor core":

"At the community level, repeated FMT failed to induce cumulative donor convergence."

In other words, repeating FMT from the same donor did not cause progressively more strains to colonize. Sometimes the second treatment provided additional improvement over the first, however additional FMTs after that were at least as likely to bring the donor and recipient farther apart again as to bring them closer.

This makes sense in light of my own experience, in which a repeat treatment from the same donor, using the same method, invariably has not resulted in meaningful additional improvement. The only time I have had meaningful effect from a second FMT from the same donor was when I changed the preparation, i.e. adding glycerol the second time when it wasn't used the first time, or changed route, i.e. used enema when I had previously used oral capsules. This has been as true for oral transplants as it has been for FMT.

What has made a difference is time from the last transplant, with changes often occurring for up to a month and a half. This clearly cannot result from completely new strains being acquired from the donor, and I'm unsure how much of it has to do with further convergence in similarity to the donor, vs. adaptation of my own existing microbiome to work together with the new microbes and "personalize" the relative levels of microbes acquired from the donor, or even gradual healing of host function after a new state of the microbiome is already established.