Key Takeaways
- The most cited cost analysis of irreproducible US preclinical research puts the figure near $28 billion a year, on an assumed 50% irreproducibility rate drawn from a literature range of 18% to 88.5%.
- In that analysis, biological reagents and reference materials was the largest weighted category at 36.1%, ahead of study design, data analysis and laboratory protocols.
- Peptides are unusually exposed to material variability. A single purity number says little about net peptide content, counterions or degradation products.
- A 2026 analysis by Eli Lilly scientists found an uncharacterized tirzepatide-B12 adduct in all 10 compounded samples tested, at up to 10% of total polypeptide content.
- Analytical documentation for the compound tested is part of a study’s evidence, not a footnote to it.
A peptide study reports an effect, while another lab runs something close to the same experiment and finds nothing. The usual explanations get reached for first: different cell line, different endpoint, small sample, publication pressure. Those are often right, but a quieter possibility gets examined last. The two labs may not have been testing the same molecule.
The Reproducibility Problem Has a Materials Component
In a survey of 1,576 researchers published in Nature in 2016, 52% agreed there was a significant reproducibility crisis, more than 70% had failed to reproduce another scientist’s experiment, and more than half had failed to reproduce their own. Respondents most often pointed to pressure to publish and selective reporting, each cited by more than 60%.
A 2015 analysis in PLOS Biology by Freedman, Cockburn and Simcoe took a different cut, putting the annual US cost of irreproducible preclinical research near $28 billion and weighting the causes across four categories. The largest portion of study was biological reagents and reference materials at 36.1%, ahead of study design at 27.6%, data analysis and reporting at 25.5%, and laboratory protocols at 10.8%.
Both figures carry a caveat: the coverage usually drops. The $28 billion rests on an assumed 50% irreproducibility rate drawn from published estimates spanning 18% to 88.5%, and the category percentages are weighted prevalence shares, not measured dollar attributions.
A 2021 comment in Nature Communications by de Marco and colleagues attributed $10.4 billion of research spending to poor quality reagents and reference materials, and proposed 3 tests as reliable indicators of quality: identity by mass spectrometry, purity by SDS-PAGE, capillary electrophoresis or reversed-phase liquid chromatography, and homogeneity by size-exclusion chromatography or dynamic light scattering. None are exotic. They are simply not always run, and the results are not always published.
Why Peptides Are Especially Exposed
Synthetic peptides carry failure modes that a single purity figure on a vial does not capture.
Purity Is Not a Single Number
A 2026 review in the Journal of Pharmaceutical Investigation by Yang and colleagues describes what identity confirmation for peptide drug substances is widely recommended to involve: at least two orthogonal techniques drawn from mass spectrometry, NMR, amino acid analysis and peptide mapping, consistent with USP-NF chapter 1503. ICH Q6A is less prescriptive, requiring specificity rather than two techniques, though it treats identification by a single chromatographic retention time as insufficient. Either way, a purity result on its own confirms that a single peak dominates the trace. It does not confirm the peak is the intended sequence.
The European Pharmacopoeia monograph for substances for pharmaceutical use shows how fine-grained this gets: peptide-related impurities are reported above 0.1%, identified above 0.5%, and qualified, meaning justified on toxicological grounds, above 1.0%. A compound advertised at 98% purity has 2% of something else in it, and whether that matters depends on what it is.
Net Peptide Content and the Counterion Question
Net peptide content is calculated by mass balance, subtracting water, counterions and non-peptide impurities from 100%. Peptides made by solid-phase synthesis commonly carry trifluoroacetate as a counterion, and it contributes real mass. A vial labeled 10 mg may hold noticeably less peptide than 10 mg, and the gap varies between suppliers and between batches from one supplier.
In a study reporting results by mass rather than verified peptide content, that gap transfers directly into the reported concentration. Two labs working from the same nominal figure can be running measurably different exposures.
Degradation Happens After Synthesis
The peptide-related impurities catalogued in the pharmaceutical literature include deamidation products, where glutamine converts to glutamic acid and asparagine to aspartic acid, oxidation products such as methionine sulfoxide and oxidized tryptophan, racemization products, terminal truncations, deletion and insertion variants, and residue from incomplete deprotection.
Some are present on day one. Others accumulate during storage, in solution and through freeze-thaw cycling. A certificate of analysis describes the material as it was when tested, not as it is after months of handling.
A 2026 Case Study in Uncharacterized Impurities
The clearest recent illustration comes from outside the research-only market, in compounded products intended for people. A 2026 paper in Expert Opinion on Drug Safety examined 10 compounded tirzepatide and vitamin B12 samples from US compounding pharmacies, medspas and telehealth networks. All five authors work for Eli Lilly and Company, which makes the branded tirzepatide products compounded versions compete with, and that conflict belongs on the table.
The analytical work holds up on its own terms. UPLC-MS on the market samples identified an adduct near 6,138 daltons, against a monoisotopic mass of 4,810.52 daltons for tirzepatide alone. Structural confirmation came from a separate step: 1D and 2D NMR on laboratory mixtures of tirzepatide and hydroxocobalamin at a 1:1 molar ratio, consistent with a B12 molecule attached through a ligand substitution reaction. All 10 samples contained the adduct, at up to 10% of total polypeptide content, and the authors called it an uncharacterized molecular species with an unknown safety profile.
The methodological lesson survives the conflict of interest. A chemically predictable impurity sat undetected across every sample tested until somebody ran orthogonal analytics looking for it. Formulation chemistry generates species nobody is looking for, and nobody finds them by reading a label.
What a Usable Documentation Trail Contains
For anyone assessing a peptide finding, as a reviewer, a clinician reading a preprint, or a researcher building on published work, the questions worth asking about the material are short and answerable.
Was identity confirmed by two independent methods, or inferred from a single chromatogram? Was purity reported with method and conditions, or as a bare percentage? Was net peptide content stated and the counterion identified? Are results traceable to a specific batch, with third-party confirmation rather than supplier assertion alone? Were storage conditions and time in storage recorded between receipt and use?
Practice varies widely across the research-supply market. Some sellers publish batch-level certificates of analysis and independent verification data, others publish a purity figure and nothing behind it, and the difference is worth checking directly against whichever research peptide supplier a study drew its material from. Journals rarely require this in a methods section, so its absence is not a mark against a paper. It signals that a variable went unreported, which is not the same as it having been controlled.
Where the Regulatory Picture Sits
Material sold for laboratory research is not manufactured, tested or labeled to medicine standards, so purity and identity claims in that market rest on the seller’s own analytical program and whatever independent testing they commission. Nothing in current US policy changes that.
The other track is compounding for human administration, where the position is moving. FDA’s Pharmacy Compounding Advisory Committee met on 23 and 24 July 2026 and recommended six of seven nominated peptides for the 503A bulk drug substances list: BPC-157, KPV, TB-500, MOTS-c, Epitalon and Semax, rejecting emideltide. The votes were close, three splitting eight to seven. As FDA states, advisory committees make non-binding recommendations, which the agency generally follows but is not legally bound to adopt, and rulemaking is not complete.
The evidence base behind those votes is the relevant part. For four of the seven, including TB-500, KPV, MOTS-c and Epitalon, FDA identified no human clinical studies for the proposed indications and no human safety studies. Where studies existed they were small, safety reporting was often incomplete, and durations ran no longer than 15 days. A recommendation advancing through a regulatory process is not an evidence base, and neither establishes what a batch contains.
Reading Peptide Research More Carefully
The reproducibility literature has spent more than two decades pointing at incentives, statistics and design. Those matter, but the largest weighted category in the most cited cost analysis is the material itself, and peptide research has more places for material to go wrong than most fields.
When a peptide result does not replicate, the molecule in the vial belongs on the suspect list alongside the experimental design, and unlike most items there it can be checked before the experiment runs.
References
1. Freedman LP, Cockburn IM, Simcoe TS. The economics of reproducibility in preclinical research. PLOS Biol. 2015;13(6):e1002165. doi:10.1371/journal.pbio.1002165
2. Baker M. 1,500 scientists lift the lid on reproducibility. Nature. 2016;533(7604):452-454. doi:10.1038/533452a
3. de Marco A, Berrow N, Lebendiker M, et al. Quality control of protein reagents for the improvement of research data reproducibility. Nat Commun. 2021;12:2795. doi:10.1038/s41467-021-23167-z [Comment]
4. Jordan B, Arbogast LW, Clemens M, Huang L, Snyder MT. A novel, widespread impurity in mass-compounded tirzepatide/B12 products: potential patient safety implications. Expert Opin Drug Saf. 2026;25(5):837-845. doi:10.1080/14740338.2026.2663185
5. Yang EJ, Kim SH, Kim A, Choi J, Jeong HJ, Na DH. Regulatory and analytical considerations for the quality assessment of peptide drugs. J Pharm Investig. 2026;55. doi:10.1007/s40005-026-00817-2
6. International Council for Harmonisation. ICH Q6A: Specifications, test procedures and acceptance criteria for new drug substances and new drug products, chemical substances. 1999.
7. US Food and Drug Administration. Meeting of the Pharmacy Compounding Advisory Committee, 23-24 July 2026. https://www.fda.gov/advisory-committees/advisory-committee-calendar/july-23-24-2026-meeting-pharmacy-compounding-advisory-committee-07232026
8. FDA advisory committee’s vote may open a drug compounding back door to unapproved peptides. Health Affairs Forefront. 2026. https://www.healthaffairs.org/content/forefront/fda-advisory-committee-s-vote-may-open-drug-compounding-back-door-unapproved-peptides
Disclaimer: This article is for educational and informational purposes only. It is not medical advice, a recommendation for treatment, or an endorsement of any product, supplier or specific use. Peptides discussed here are research compounds. They are not approved by the FDA for therapeutic use and are not intended for human or animal consumption. Anyone considering a peptide product for health purposes should consult a licensed physician.
