The perils of a vanity press in scientific publishing

The perils of a vanity press in scientific publishing

This section details the interactions of IROA Technologies LLC with the authors of  metabo12080741 and the editors of the journal Metabolites and their parent organization.  The basis for the original letter to the editor was accepted and agreed to by the author in their first reply.  The authors demonstrated that all of the points made in the letter to the editor were correct but MDPI had accepted a cash payment to publish the article in question and declined to remove it.  This prompted a second letter to the editor that noted the scientific basis for the publication had not only been accepted by the authors although the article offered no scientific value or merit.  The authors, who had an editorial relationship with the journal as well as a commercial relationship, again substantiated the second letter to the editor.  The editor admitted the problems with the article, however instead of  requesting  the retraction that it deserved they instead asked for a revision .

The relevant timeline

 On August 12, 2022 the journal Metabolites (from MDPI, LLC) published an article, metabo12080741, that claimed to have used the IROA TruQuant protocol and IROA materials to determine if the protocol could be used for urine metabolomics.  Upon close inspection of the paper four things were clear 1) that the authors had not used the published protocol, but rather made one up, 2) that the samples used had been overly dilute because the authors had failed to balance the amount of sample that was used with the internal standard, and 3) that the authors had calculated their results using a method that introduced very significant (20% to 60%) error into their data, and therefore the statics based on this erroneous data was suspect, and 4) since they had not followed the protocol the main benefits of the protocol, the error-correcting steps of ion-suppression correction, and sample-to-sample normalization, were unavailable to them for analysis.   On August 28th (16 days post publication) we notified the the editor in our "First letter to the editor" (Available here).  

The Authors replied in their "First authors rebuttal" (Available here) on September 18, 2022 (37 days post publication).  In their reply the authors agreed that 1) they had not followed the protocol but had adapted it in an acceptable way to suit their needs , 2) they had not calibrated the sample concentration to the internal standard, but post receipt of our letter they did calibrate the samples and had found that their samples were between 1% to 50% of the concentration needed, 3) they stated but did not support that their method of calculating the C12/C13 (natural abundance to internal standard) ratio using only the values of the monoisotopic peak was correct and not erroneous, and 4) their ratio values were the equivalent of a suppression corrected value.  The editors declined correcting or retracting the paper despite 1) our protocol being named, as if used, in the title, abstract and throughout the paper, and 2) the admission that the totality of the data was based on samples that were extremely dilute and in some cases approaching LLOQ (on the sample side).

MDPI Ethical Guidelines for Authors

• Communicate errors and inaccuracies found after publication.

• Present their data and methods with attention to detail. Data and methods used in the research need to be presented in sufficient detail in the manuscript so that other researchers can replicate their work. Raw data must be made publicly available unless there is a compelling reason otherwise (e.g., patient confidentiality) 

We replied to the editors ("Second Letter to Editor"  available here) asking how the editors could justify neither correcting nor retracting a paper in which the authors admitted to two clear wrongs 1) they claimed use of a protocol that they had not used, and 2) used samples that were so dilute they had caused one of the outcomes of the authors  on November 9, 2022 (89 days post publication).  In this second letter we also delineated why the authors' method of calculation was wrong, reiterated that all of the statistical outcomes of erroneous data had to be understood to contain error when it was fairly easy to correct the error and ultimately the statistics.  IROA TruQuant protocol is based on two error-reducing techniques, the correction for in-source losses (ion-suppression correction) and the correction for sample-to-sample variances (sample-to-sample normalization).  With this second letter we also questioned why the authors did not include these outcomes in their analysis, and formally asked the editor to help us acquire the raw data so we could process it ourselves.  We had previously asked the authors for the raw data but they had never responded.  

The authors reply ("Second author's rebuttal" available here) was received, from the editors, on January 28th 2023  almost 6 months after publication.  In this second reply the authors make several new clear mistakes.  In defending their alterations of the protocol they cite many IROA publications none of which used the IROA TruQuant protocol including one of our own papers,  Beecher C; de Jong FA. (2019): Isotopic Ratio Outlier Analysis (IROA) for Quantitative Analysis: Springer New York (Methods in Molecular Biology).