PubChem is a free chemistry database run by the US National Center for Biotechnology Information. It is one of the first places researchers look up a compound, and links to PubChem records are common in papers, supplier pages and forums. Reading a PubChem record well means understanding how the database is built. Many confusions about peptides come from treating a PubChem record as if it were a description of a particular vial.
This guide explains the structure of PubChem, the identifiers you will see, and the pitfalls that matter for peptide research.
How PubChem Is Organised
A 2016 paper by the PubChem team in Nucleic Acids Research describes three inter-linked databases:
- Substance (identifiers called SIDs) holds depositor-provided descriptions of chemical substances. If ten organisations each deposit a record for the same molecule, there are ten substance records.
- Compound (identifiers called CIDs) holds unique chemical structures extracted from the Substance database through a process called standardisation. Substance records about the same molecule from different sources are aggregated through a common compound record.
- BioAssay (identifiers called AIDs) holds depositor-provided descriptions and results of biological experiments on substances.
The scale is large. The PubChem 2025 update in Nucleic Acids Research reports more than 1,000 data sources, 119 million compounds, 322 million substances and 295 million bioactivities.
Why this matters
A CID describes a structure, not a product, purity, salt form or lot. A SID describes what one depositor submitted, and different depositors may describe the "same" compound differently. A BioAssay result describes one experiment with one sample under one set of conditions.
Standardisation: How Many Records Become One
The 2016 paper describes how PubChem standardises deposited structures:
- It checks that a structure is chemically reasonable.
- It normalises equivalent or alternative valence-bond structures, meaning tautomeric and resonance forms, into a single representation.
- It annotates stereocentres where their configuration is defined.
- If the standardised structure is a mixture with multiple covalent units, the separate covalent structures are isolated and processed individually.
- Substances are mapped to compounds using structural hash codes calculated from the standardised structure.
Two consequences matter for peptides:
- Salts and counter-ions are split. A peptide deposited as an acetate or trifluoroacetate salt is a multi-component substance. Standardisation separates the peptide from its counter-ion, so the peptide's CID may not describe the material as supplied. Molecular weights shown on a compound record may be for the free peptide, not the salt. See Net Peptide Content vs. HPLC Purity.
- Stereochemistry matters. Peptides built from L-amino acids have defined stereocentres. A record with undefined stereochemistry, or with a D-amino acid, is a different structure and can carry a different CID. Check that the record matches the intended sequence and configuration.
Why Many Peptides Have No 3D Conformer
PubChem provides computed 3D conformer models, but only for compounds that meet specific conditions. The 2016 paper lists them:
- ≤50 non-hydrogen atoms and ≤15 rotatable bonds (not too large, not too flexible),
- a single covalent unit (not a salt or mixture),
- only supported elements,
- only atom types supported by the MMFF94s force field, and
- five or fewer undefined stereocentres.
About 90% of compounds in PubChem meet these conditions. Most peptides longer than a few residues do not, because they exceed the size or flexibility limits. A missing 3D structure on a peptide record is normal. It does not mean the record is incomplete or wrong.
CAS Registry Numbers
Alongside CIDs and SIDs, many records list a CAS Registry Number (CAS RN). According to CAS, a division of the American Chemical Society:
- A CAS RN is a numeric identifier of up to 10 digits, divided by hyphens into three parts.
- The right-most digit is a check digit used to verify the validity of the whole number.
- CAS REGISTRY contains 300+ million substances. CAS Common Chemistry, an open resource, offers nearly 500,000 of them.
The check digit gives you a quick way to catch typos. Take every digit except the last, multiply each by its position counted from the right (1, 2, 3...), add the products, and take the last digit of the sum. For water, 7732-18-5: 8×1 + 1×2 + 2×3 + 3×4 + 7×5 + 7×6 = 105, so the check digit is 5. A CAS number that fails this check has been copied wrongly.
A valid check digit does not prove the number belongs to your compound. It only proves the number is well-formed. Different salt forms, hydrates and stereoisomers can have different CAS numbers, and some peptides have several CAS numbers in circulation.
Synonyms: Useful and Risky
PubChem compound records gather synonyms from all depositors: systematic names, trade names, codes and abbreviations. That makes searching easy. It also brings risk:
- Synonyms are depositor-supplied. A name on the synonym list was submitted by someone. It is not a verified assignment.
- Abbreviations collide. Short peptide names and codes can point to different molecules in different literatures. See What Peptide Names Mean and the TB-500 example in TB-500 and Thymosin Beta-4.
- Fragments and analogues share roots. A name search may return the full peptide, a fragment, an acetylated form or a salt, each with its own CID.
The safest key is the structure itself: the sequence, terminal modifications and stereochemistry, or a structure identifier such as an InChIKey.
BioAssay and Literature Data: Read With Care
Compound records link to bioactivity results and literature. The 2025 update describes new interfaces such as a consolidated literature panel and a patent knowledge panel. These links are a starting point for reading, not a summary of evidence:
- A bioactivity result is depositor-reported, from one assay with one sample.
- A literature link shows that a paper mentions the compound, not what the paper concluded or whether it holds up. See When a Paper Is Retracted.
- An assay in cells or tissue is a different model from an animal or human study. See Dish, Mouse or Human?.
A Checklist for Matching a Record to a Vial
- Sequence and modifications: Does the structure on the record match the expected sequence, terminal groups (for example amide vs. free acid) and any cyclisation or disulfide?
- Stereochemistry: Are configurations defined and correct?
- Salt form: Is the record for the free peptide, while the vial contains a salt? Expect a different molecular weight.
- CID vs. SID: Are you looking at the standardised compound or one depositor's substance?
- CAS RN: Does it pass the check digit, and does it correspond to the same form?
- Molecular formula and mass: Do they match the identity data on the analytical report? See How to Read a Peptide Analytical Report.
- Synonyms: Use them as search aids, not proof of identity.
FAQ
Is a PubChem record proof of a compound's identity or quality? No. It describes a structure and links depositor data. Identity and purity of a specific batch come from analysis of that batch.
Why does the molecular weight on PubChem differ from the label? Often because PubChem shows the standardised free peptide, while a label may refer to a salt form, or vice versa.
Why are there several CIDs for what seems to be the same peptide? Differences in stereochemistry, terminal groups, salt separation or modifications produce different standardised structures.
Sources
- Kim S, Thiessen PA, Bolton EE, et al. PubChem Substance and Compound databases. Nucleic Acids Res. 2016;44(D1):D1202-D1213. https://pmc.ncbi.nlm.nih.gov/articles/PMC4702940/
- Kim S, Chen J, Cheng T, et al. PubChem 2025 update. Nucleic Acids Res. 2025;53(D1):D1516-D1525. https://pmc.ncbi.nlm.nih.gov/articles/PMC11701573/
- CAS. CAS REGISTRY and CAS Registry Numbers: https://www.cas.org/cas-data/cas-registry
Every Peptide Privé batch is tested by an independent lab.
Contact our team with documentation questions.
For in-vitro laboratory research only. Not for human or animal use.
