Published October 12, 2026 | Version 2026-10-12

TitrateLab peptide COA analysis: vendor-blind aggregates of 96,805 published lab reports (release 2026-10-12)

Authors/Creators

Description

What this is. Monthly, vendor-blind aggregates of published peptide certificates of analysis (COAs), the lab reports that vendors and buyers post publicly. Release 2026-10-12 covers 96,805 included reports across 68 canonical peptides and blends, 26 named laboratories and 45 observed calendar months (2022-07 to partial 2026-10; 44 completed months through 2026-09), drawn from 116,363 published entries (115,076 distinct certificate identifiers, an upper bound).

Tables. compounds.csv (one row per compound), compound_month.csv, labs.csv (one row per laboratory label, with a profile link where one exists), lab_month.csv and months.csv, plus manifest.json with coverage, missing-data counts, laboratory rules and SHA-256 hashes, and compounds.json included in the versioned release. Columns are defined in schema/data-dictionary.md inside the archive.

Headline figures in this release. 49.9% of 68,267 quantity results fall within 10% of the printed label; 12.9% of included reports carry an endotoxin result; 671 of 87,265 positive-or-zero purity results (0.77%) report zero purity or an unidentified target compound; 98.4% carry a laboratory verification URL.

Read the denominators. Rows are aggregates of published lab-report entries, not individual reports or unique batches. Quantity-within-label percentages use the 68,267 entries with a quantity-versus-label result; 28,538 included entries have none. Purity means and medians use the 86,594 positive-purity results and exclude the 671 zero-purity/unidentified-target entries. Endotoxin-result coverage uses all 96,805 included entries: 12,516 carry a result, including below-detection outcomes, and 84,289 do not. Missing measurements are neither zero values nor passing results. These are coverage and measurement summaries, not market prevalence or product safety estimates.

Privacy boundary. No vendor, manufacturer, client, submitter, batch or certificate identity appears in the released aggregate data tables. Every cell holds at least 10 records; smaller cells are omitted and small laboratories fold into an 'Other labs' row so totals reconcile. Laboratories are named because they are the businesses that issue and publish the reports.

Limitations. The corpus is not a random sample of the market: vials are tested and published by choice, and collection coverage differs by laboratory. Quantity deviation is measured against the printed label only. Extraction is imperfect; completeness is published per field in the manifest rather than assumed. Nothing here is medical advice or a statement about the safety of any product.

Method and provenance. Built from TitrateLab's read-only projection of published lab reports (projection 20261012T021040Z, records observed through 2026-10-12T00:23:06Z). Collection, deduplication, canonical-name resolution and quality rules are documented in the methodology/ directory of the archive. Every number in the README and analyses is rendered from these files by scripts/build_release.py. Per-record evidence with laboratory, client and batch as printed is searchable in the public lab-report index.

License and reuse. Data and analysis: CC BY 4.0; original code: MIT. Attribute TitrateLab, cite this dated release, link the license and state your changes. Include the metric denominator and limitations beside any reused chart or calculation. This archive is the October 12, 2026 snapshot; current website totals may differ.

Suggested citation. TitrateLab. (2026). TitrateLab peptide COA analysis: vendor-blind aggregates of 96,805 published lab reports (release 2026-10-12) (Version 2026-10-12) [Data set]. Zenodo. https://doi.org/10.5281/zenodo.23551719.

Find release context in the TitrateLab dataset catalog.

Read the measurement and collection methodology before interpreting a result.

Reproduce the analysis from the fixed GitHub source release; column definitions are in schema/data-dictionary.md inside the archive.

Files

peptide-coa-analysis-v2026-10-12.zip

Files (830.1 kB)

Name Size
md5:efd948eb0b7d8aa646533851a1a56338
830.1 kB Preview Download

Additional details

Related works