There are serious issues with the paper titled, “16S rRNA gene sequencing and healthy reference ranges for 28 clinically relevant microbial taxa from the human gut microbiome” (DOI: 10.1371/journal.pone.0176555), authored by Almonacid et al., published in PLOS ONE in 2017. While the Expression of Concern issued on October 20, 2022, acknowledges some of these problems, it’s necessary for the authors to fully address them due to the broader implications for scientific integrity.
I previously attempted to raise these significant concerns on PubPeer on November 19, 2024; however, my comments were completely censored by the moderator and never published.
1- Questionable Data Integrity: The Expression of Concern has already noted that one sample (sample 757) may have originated from a non-human source, raising questions about data verification practices. The presence of potentially non-human data in a study purporting to define “healthy human” microbiome reference ranges compromises the validity of its conclusions and indicates potential systematic oversights that could affect other data points.
2- Reliance on Self-Reported Participant Health: The reliance on self-reported data without independent medical verification is a significant weakness, especially in a study establishing baseline microbiome health metrics. This methodology introduces a substantial risk of undetected medical conditions among participants, potentially skewing results and reducing the credibility of the reported healthy reference ranges.
Looking forward to a constructive discussion.
I thank the SG Participant for their comments. Both comment have been acknowledged in the Expression of Concern, so I am not sure which new concerns the Participant is trying to raise.
Ad 1. One of the samples indeed may have been of non-human origin, as already mentioned in the EoC. The emphasis is on may, as one customer sent in multiple samples, and they were not clearly labeled. While we might regret that one out of nearly 900 samples might not have been carefully selected, there might be reasons to assume the correct, human, sample was included in the study. Looking at the microbiome composition of this particular sample, it is not an outlier at all, falling within the ranges of the human samples. So it did not skew the results in my opinion.
Ad 2. The study is clearly marked as relying on self-reported health data. In that respect, it is not unique. Many studies use questionnaires in which participants self-report their feelings, income, weight, health, and other data. Even though researchers know those answers might not be completely reliable, they would still publish papers based on those questionnaires. With large datasets like the one reported in this paper (nearly 900 samples), we assume data might have been reported correctly – that is an assumption that is not unique for this study.
Both of these points have been raised and answered in the EoC – which is why this discussion feels a bit old. Were there any new issues the SG Participant wanted to raise or did I miss them?