The SS-31 dispute is no longer just about one peptide, one report, or one lab result. It’s turned into something much messier and more personal. A public fight over methods, machines, raw data, AI accusations, subcontracted analysis, laboratory transparency, scientific competence, and who the grey-market community is supposed to trust when two labs produce completely different conclusions.
I covered the original FSD SS-31 disagreement in the article linked below.
Battle of the Labs: The Equipment Edition
The first article focused on the initial split, how UZORAK reported hidden impurities and PVDF contamination, while Janoshik later said he specifically looked for those issues and couldn’t reproduce them. This article picks up from there. Because what followed was no longer just a disagreement over SS-31. It became a fight over whether the methods themselves were valid, and then, very quickly, over the people behind them.
The Dispute Moves Beyond “Better Machines”
What followed was no longer simply an argument about whether one lab had a “better machine.” The dispute escalated into a much broader fight over chromatographic setup, sample preparation, co-ionization, trapping-column usage, calibration, quantitative interpretation, raw-data analysis, and ultimately whether the conclusions themselves were being interpreted correctly.
The Deamidation Argument Evolves
One of the major technical disputes involved deamidation, which produces a mass shift of approximately +0.98 Da. Because that sits very close to the natural +1 isotope pattern, this became an argument over what a lower-resolution system should still be able to see.
Janoshik’s point was simple, that even if a single quadrupole cannot perfectly separate +0.98 Da from a natural +1 isotope peak, a meaningful amount of deamidation should still change the overall isotope distribution. In other words, it may not resolve perfectly, but it should still disturb the pattern.
UZORAK’s counterpoint was that this depends on the workflow around the MS data, i.e. chromatographic separation, salt interference, co-ionization, trapping-column clean up and interpretation. So the question became less “Can a single quad technically see it?” and more “Was this specific method set up well enough to detect it or rule it out reliably?”
But later another technical criticism emerged from Janoshik around the deamidation claim itself. He pointed out that SS-31 does not contain the classical deamidation-prone amino acid residues normally associated with common peptide deamidation pathways, such as asparagine (Asn) or glutamine (Gln). Instead, the only theoretically relevant site would involve slow hydrolysis/deamidation of the C-terminal amide under harsher conditions. His argument was essentially that while deamidation isn’t chemically impossible, the peptide sequence itself doesn’t strongly support the level of deamidation being claimed.
The Acetylation Fight
Acetylation became another major point of conflict. Unlike deamidation, acetylation involves a much larger mass shift approximately +42 Da. Janoshik argued this should be straightforward to detect even on single quadrupole systems if present at meaningful levels, but the argument later evolved into a dispute over quantitation itself.
UZORAK publicly suggested that screenshots from Janoshik’s own chromatograms mathematically appeared closer to ~1.5% acetylation rather than the reported <0.1%. Janoshik pushed back aggressively against this interpretation and accused UZORAK of using AI to interpret this. Janoshik’s criticism was that UZORAK appeared to be deriving that estimate from visible peak height in screenshots, rather than from integrated area under the curve. In chromatography, quantitation is normally based on AUC, not raw peak height, especially in this case, where Janoshik intentionally overloaded the sample in an attempt to force low-level impurity detection, which distorted the peak shapes and pushed the detector outside ideal linear conditions.
So this became a fair criticism of UZORAK’s public interpretation, not necessarily proof that Janoshik’s result was correct, but proof that screenshot-based maths was not enough.
The Acetate Salt Question
As the discussion continued, Janoshik’s side expanded the argument further. He pointed out that SS-31 is commonly manufactured as an acetate salt, meaning trace acetylation-related signals may not automatically indicate severe impurity or degradation. He also argued that under harsh ionization or source conditions inside the mass spectrometer, acetate-related interactions and artefacts could potentially contribute to low-level acetylated species appearing during analysis itself. He said that if SS-31 truly contained the level of acetylation being claimed by UZORAK, the polarity change should have been much more obvious on routine HPLC analysis.
The “Mannitol vs PVDF” Escalation
Earlier, the most unresolved part of the entire dispute was the PVDF contamination only detected by UZORAK. Unlike peptide degradation products, PVDF is not a simple mass shift, a peptide variant, or a normal degradation product. It’s a fluoropolymer contamination claim. That changes the entire analytical discussion.
UZORAK claimed PVDF-related signals were detected across multiple SS-31 samples from multiple vendors, including FSD, ERP and HYB products. Initially, the argument evolved into a much deeper discussion involving trapping columns, polymer ionization, solvent behaviour, chromatographic clean up, co-ionization, and sample preparation itself. And one important issue that was raised was that PVDF itself is poorly soluble in water. That then led to competing arguments about whether literal PVDF particles could realistically pass through certain workflows, whether dissolved or degraded fluoropolymer-related material was instead being detected, whether solvent systems such as acetonitrile gradients could partially mobilize polymer-related species, and whether trapping-column clean up was necessary to separate those signals from salts and formulation components.
But the most explosive escalation came when Janoshik publicly argued that UZORAK’s alleged PVDF signal may not represent PVDF at all. Instead, he claimed the spectra looked more consistent with mannitol-related clusters. That accusation dramatically raised the stakes since mannitol is not an exotic contaminant. It’s just a common excipient used in lyophilised peptide formulations as a bulking agent, stabiliser and lyoprotectant.
So if the signal was actually mannitol-related chemistry rather than fluoropolymer contamination, it would fundamentally change the meaning of one of UZORAK’s most alarming findings. Mannitol has the formula C₆H₁₄O₆ and a molecular mass around 182 Da, which makes Janoshik’s interpretation more plausible, that the disputed signal may fit a common lyophilised peptide excipient better than PVDF contamination.
Janoshik then went on to suggest that UZORAK’S AI-assisted interpretation may have contributed to the misidentification. His claim was that screenshot-based, AI-assisted spectral interpretation may have incorrectly associated the observed pattern with PVDF, instead of recognising a more ordinary filler-related signal.
At that point, the dispute deteriorated rapidly into accusations of AI-generated science, analytical incompetence, weak chemistry background, subcontracted work, and personal credibility attacks.
But underneath the theatre, the scientific question remained serious.
If Janoshik is wrong, then UZORAK may have identified a genuine polymer contamination issue that Janoshik’s workflow failed to reproduce, but if Janoshik is right, then one of the most alarming contamination warnings in the peptide-testing world may have come from misidentifying excipient-related spectra as fluoropolymer contamination.
Raw Data vs Reality
One of Janoshik’s strongest counterarguments was that raw data alone can’t resolve everything. Raw instrument files can show what the machine recorded, but they can’t automatically explain cross-contamination, sample mix-ups, handling differences, preparation differences, or whether two labs truly analysed chemically identical material. That’s an important point because even perfect machine files still sit inside a wider chain of sample preparation, operator decisions, chromatography conditions, dilution choices, handling procedures and uncontrolled grey-market supply chains
Raw data is extremely valuable, but it’s not magic. It can’t retroactively guarantee chain-of-custody, sample integrity, preparation consistency, or identical source material. And that may be one of the biggest lessons from the entire SS-31 situation.
When the Science Became Personal
As the debate escalated, the tone deteriorated rapidly. Janoshik openly accused Alex/UZORAK of relying heavily on AI-generated scientific arguments, misunderstanding chromatography, and making analytically invalid conclusions from screenshots. He also questioned Alex’s chemistry background, whether UZORAK functioned primarily as a middleman, and whether some analytical work was subcontracted through outside institutions.
At the same time, UZORAK aggressively attacked Janoshik’s method setup, interpretation, calibration, chromatographic workflows, and overall analytical competence.
By this stage, the dispute had turned into “who should the community trust to interpret what the machine detected?” And once a scientific dispute reaches that point, then personality, transparency, communication style, reputation, and perceived credibility, inevitably become part of the story.
Importantly though, even while attacking Alex personally, Janoshik repeatedly clarified that he was not necessarily claiming that the underlying data itself was fabricated. Instead, his criticism focused more on methodology, interpretation, presentation and competence.
The Ruđer Bošković Question
The Ruđer Bošković Institute entered the discussion after Janoshik began questioning Alex’s chemistry background, whether UZORAK was acting mainly as a middleman, and whether some of the analytical work was being performed through this institute, rather than entirely in-house. In response, supporters of UZORAK pointed out that the institute is not some unknown back-room operation, it’s Croatia’s largest multidisciplinary scientific research centre, with established work across chemistry, molecular biology, physics and analytical science. While a respected institute can produce serious analytical data, the interpretation, presentation, commercial packaging, or downstream conclusions attached to that data can still remain open to debate. So the institute discussion doesn’t prove UZORAK correct, but it does change the nature of the criticism.
The Wider Problem This Exposes
This whole SS-31 saga shows how complicated grey-market peptide testing can be. But it’s also exposed something else, the rivalries between labs, the politics around interpretation, and how quickly scientific disputes become public reputation wars once trust itself is on the line.
It’s important to note that UZORAK had left the thread early and was absent for most of this discussion, so not able to defend each point as it developed. That matters, because public debates are shaped by who stays in the room, so while Janoshik’s side gained momentum, especially around the mannitol argument, that doesn’t mean every question is fully settled.
I will also be posting UZORAK’s defence and any further analysis that comes from the raw data.
Comments