A sequence search can locate records that a name search misses, but its meaning depends on the tool. An exact substring match, a local similarity alignment and a text search for residue letters answer different questions. None automatically proves that a published experiment used the same chemically modified peptide.
Define what the sequence does and does not encode
Start with a checked residue sequence in the notation the tool accepts. Keep a separate identity note for terminal groups, stereochemistry, cyclisation, conjugation or other features that an ordinary one-letter string may not represent.
For an original fictional example, the string AGLKSTV describes seven residue letters. It does not by itself distinguish a free terminal group from an amidated one, nor does it document the provenance of the sequence.
| Check | Reason |
|---|---|
| Residue order | A transcription error changes the query |
| Full length or fragment | A fragment match has narrower meaning |
| Nonstandard chemistry | Plain letters may omit identity features |
| Database scope | The target record may not be included |
| Taxonomy restriction | An unnecessary restriction can remove matches |
These checks concern retrieval. They are not instructions for synthesising or preparing a peptide.
Read an exact peptide match correctly
EMBL-EBI’s UniProt training describes peptide search as finding matching sequences within UniProtKB records. It also describes an option to treat isoleucine and leucine as equivalent, which changes the interpretation of a match.EMBL-EBI — UniProt peptide search (opens in a new tab)
In the fictional example, finding AGLKSTV inside a much longer protein establishes a matching segment under the chosen settings. It does not show that the isolated seven-residue material was experimentally studied in that record.
If several proteins contain the same short segment, a match cannot identify one unique source protein without additional evidence. Record the matched positions, organism and accession rather than retaining only the top result.
Inspect local alignments and query coverage
NCBI’s BLAST documentation identifies blastp-short as a task optimised for protein queries shorter than 30 residues. Short-query settings matter because a short peptide and a long protein do not provide the same amount of sequence information.NCBI — BLAST command-line applications, blastp options (opens in a new tab)
A local alignment can cover only part of the query. An illustrative ten-residue alignment showing 100% identity across five aligned residues does not establish a full ten-residue exact match; the remaining query positions still need examination.
Read aligned length, query coverage, mismatches and gaps together. A high identity percentage without its denominator can conceal how little of the query was compared.
Do not treat a similarity score as proof of biological function. The search identifies a sequence relationship under its model and database, while functional evidence requires the relevant experiments and annotations.
Follow sequence records into the underlying research
Use a matched accession and its linked publications to locate the studies behind the record. Then check whether the paper concerns the full protein, a processed fragment, a synthetic analogue or a different experimental material.
A sequence database can help locate the parent context without containing every chemically modified research peptide as its own entry. Failure to retrieve an exact record therefore does not prove that a compound or study does not exist.
Keep the tool, database, run date, query and relevant settings with the results. Interface limits can differ between web and programmatic services, so consult the current service documentation instead of treating one historic input limit as universal.
Finish with a bounded conclusion such as a matching segment was located in these records, followed by the identity features still unverified. That is more informative than labelling every sequence hit as evidence for the supplied product.
Sources and further detail
- EMBL-EBI — UniProt peptide search (opens in a new tab)
Exact-match purpose and isoleucine/leucine option read. Web training and API documentation give different minimum lengths; no universal minimum is prescribed here.
- NCBI — BLAST command-line applications, blastp options (opens in a new tab)
Task descriptions read, including blastp-short. No actual sequence search, score, accession or biological finding is invented.
Sources checked 19 September 2026. Worked examples are illustrative unless a supplied report is explicitly identified. This article has not undergone independent scientific peer review.