# Demo data

Version 1.1.0, 2026-11-15.

Small files for trying each part of Sequences, made only from published patent documents and public database
entries. On the search page, each **Use demo data** button fills one input with one of these files, and in a case's
**Report** step **Use a demo CSV** fills the result file with `blastp-demo.csv`. Each file is also served for
download at `/demo/<file>`, open without sign-in, so you can save it and upload it yourself.

Nothing here is made up. Every sequence, claim and result comes from the source named below, and
`scripts/sequences-demo-data.mjs` 1.0.0 made every file from those sources, except `blastp-demo.csv` and its
queries, which `scripts/sequences-column-map-files.mjs` made. A file whose SHA-256 differs from the table below is not
the demo file.

## Files

| File | What it holds | Where to use it | Bytes | SHA-256 |
| --- | --- | --- | --- | --- |
| `us12788009b2.xml` | US 12788009 B2, the grant's full text as the US patent office published it: its 9 claims and description | **Publication XML file** in a case's **Claims** step | 220,869 | `d11ef5ae9987271855c1ddd30a0af278d9b0bdd4547a66d9116635a8d73abf49` |
| `us12788009b2-claims.txt` | The same 9 claims as text | **Paste the claims** or **Word or text file** in a case's **Claims** step | 1,903 | `c628f11cb397cfa9b00688b8fbe1dd566e660815058651caff5c8d709c18bab0` |
| `us12788009b2-listing.xml` | US 12788009 B2's sequence listing, ST.26 as the US patent office publishes it, inside a `sequence-cwu` document: 16 peptides | **Listings** import; the listing of a demo case | 20,473 | `499fc6191a9acc68398e7b1f2b0fc6eddcbd7fc801f9d07785183dd95b2a464c` |
| `us12788009b2-peptides.fasta` | The listing's 16 peptides, FASTA | A protein query, on one set or several | 1,339 | `6ac400e7c5f421f1c0a0c51355e88aa3fb7934094b6f5a7898b703fb67cd14b9` |
| `us12788009b2-peptides.pir` | The same 16 peptides, PIR (NBRF) | A protein query from a PIR file | 1,403 | `d1aeb14ff9098cd4fe5cece04b2f2d074baa3cd9102f1686439f2c2510867f18` |
| `us12788009b2-seq1.fasta` | SEQ ID NO: 1, AIMARPQVPRL, which claim 1 recites | **Exact text** with a protein query | 85 | `38c5f584a5d399c2267cf48e380b926893221dcd50d4fd73b59c2f38d1e764c6` |
| `us12788009b2-pattern.txt` | Claim 1's peptide as a PROSITE pattern, `<x(0,2)-AIMARPQVPRL>` | **Pattern**, PROSITE, protein | 21 | `6219cc1cbb0cf6d4527e200ff28b4168cc810cd948ca40251174cc2a78d2507f` |
| `us12788009b2-pattern-regex.txt` | The same pattern as a regular expression, `^.{0,2}AIMARPQVPRL$` | **Pattern**, regular expression, protein | 20 | `6519263daa39bf0bd910bb58a6472dd37eb3f6e7ef75a86c67e4d8f59fc49324` |
| `us12788009b2-blastp.tsv` | BLAST+ tabular output (`-outfmt 6`): the 16 peptides against `s1-patent-proteins.fasta` | **Import a result file** in a case's **Report** step | 3,701 | `615d253b163c3b2723d4a035bd8cc7a9c242a2c48ccc04ef7a899972c1eaacb9` |
| `blastp-demo.csv` | BLAST+ output for three published proteins against 59 public protein records, saved as CSV with BLAST+'s own column names as its header row | **Use a demo CSV** under **Import a result file** in a case's **Report** step, which opens **Match the columns** | 1,023 | `0c721dcee26f6966647a83aa514d5c8367ed4377e6495d4fb43903119afa235e` |
| `blastp-demo-queries.fasta` | The three proteins of that run | A protein query | 1,061 | `9054bd43ef5ff20f9e9e9b47b4fa07deee5ee12ad383709d121d44ff0463a465` |
| `us12788762b2-primers.fasta` | US 12788762 B2, SEQ ID NO: 1 to 3: a primer pair and its probe, DNA | A DNA query | 285 | `385ed38a1a8cdbe147c3b2e7769645aa343a7e07ccf4f56ff6d79dfd5e067ca0` |
| `us12788762b2-probe.fasta` | SEQ ID NO: 3, the probe | **Exact text** with a DNA query | 95 | `4ae36d163de46c11a31bbd242a8c0dc1204320c4a4bac991f928f0bf1c9fcfd2` |
| `us12788762b2-primer-pattern.txt` | SEQ ID NO: 1, a primer with the IUPAC codes R and S | **Pattern**, either syntax, DNA | 27 | `617e492030000e83a8a2fb7a58b821099fec245e16bb6eced032437844345e7f` |
| `s1-patent-proteins.fasta` | The 26 protein sequences of the three grants' listings, each named as Sequences names a patent sequence, such as `US12788009B2_1` | A database of your own (**Import a database** on the local workbench); the database of the BLAST+ run | 3,465 | `87f7da3f560f80230a8cf320b29927aeecf0c7de730d36be11a6ba66418769ec` |
| `trastuzumab-chains.fasta` | Trastuzumab's heavy and light variable domains, from PDB 1N8Z | **Antibody CDRs**, **CDRs of antibody chains** | 261 | `1dbded3afd3eef888be16f40030b8698a4e98bac52bf9b46e0b3b06792b5e091` |
| `insulin-cds.fasta` | The coding sequence of human preproinsulin, NCBI RefSeq NM_000207.3, bases 60 to 392 | **Sequence tools**, **Editor** (check, translate, six frames) | 415 | `166f8750b3512ff05cc9c6cda6b5cc924008deacef168468281afb7b10035fb5` |
| `us12421493b2-st26.xml` | An ST.26 file as the applicant filed it: the first 3 sequences of US 12421493 B2's listing | **Check a listing before you file it**, which it passes; **Listings** import | 6,076 | `ec2680704fbbd1f57184764cef811d4c1ab06e45028f6173603ef8d3e0214933` |
| `us20260345831a1-with-listing.xml` | US 2026/0345831 A1, a publication whose full-text XML holds its ST.26 listing | **Import a publication's full-text XML** under **Listings**; a file dropped on the search card | 16,031 | `06ddfb3663efae4cd911d4f4a820448a860458460e70b08a2b83f82c9dbb6c92` |
| `api-batch.json` | A body for `POST /api/batches`: SEQ ID NO: 1 of US 12788009 B2 and the probe of US 12788762 B2, each by exact text | A program with an API token. Add `"snapshotId"` from `GET /api/databases` to each search first | 503 | `51907bd351f11ace323c1da260640f5f5c9386bc36fab91942c0d2afe714833b` |
| `connector-manifest.json` | A manifest for the document connector: the grant's XML and its listing, each with its SHA-256 | A system with a connector token: save it as `manifest.json` beside `us12788009b2.xml` and `us12788009b2-listing.xml`, and give the account the case is for as `owner` | 494 | `c2f74e4a789718db44f142f3d5bd1db2a0decb3a1bc0301c0e4fc2d50a8ea3c9` |

Two demo inputs are words, not files:

- **Watch an applicant**: `Solasta Bio Limited`, the applicant US 12788009 B2 names first.
- **API tokens**: the token name `Demo program`.

## Sources

- **US 12788009 B2**, "Insect neuropeptides 2", granted 2026-11-10 on application US 19/044,923, filed 2025-02-04.
  Its full text (document 499 of the weekly file) and its listing, `US12788009-20261110-S00001.XML` (document 500),
  are byte for byte as the US patent office's weekly grant full-text file `ipg261110.zip` holds them (198,741,802
  bytes, SHA-256 `8b6a391cc29a0588d1a8e2a0c9b56efebc13922df595edf8db5ec80c892e23eb`). Claims 1 and 5 recite SEQ ID
  NO: 1 and 2.
- **US 12788762 B2**, "Azelastine as antiviral treatment", granted 2026-11-10. Its listing,
  `US12788762-20261110-S00001.XML` (document 1,254 of the same weekly file), is ST.25 in XML elements. SEQ ID NO: 1
  to 3 are the primers and probe its description names; SEQ ID NO: 4, the SARS-CoV-2 spike protein, is the one its
  claims recite.
- **US 12788700 B2**, "Hair cleanser composition, methods and uses thereof", granted 2026-11-10. Its ST.26 listing,
  `US12788700-20261110-S00001.XML` (document 1,188 of the same weekly file), has 1,131 entries; the 9 that hold
  residues are SEQ ID NO: 75, 400, 401, 402, 409, 412, 432, 951 and 1131, which claim 14 recites.
- **US 12421493 B2**, granted 2025-09-23: its ST.26 file as the applicant filed it, from the US patent office's
  sequence data service, cut after SEQ ID NO: 3 of 818,254, with `SequenceTotalQuantity` changed to 3 to match.
  Every line kept is as the source wrote it. The listing the office publishes inside its weekly files drops the
  applicant's header, which the check needs, so a file as filed is the one to check.
- **US 2026/0345831 A1**, published 2026-11-05, from the US patent office's weekly application file
  `ipa261105.zip`: the application without its drawings and description, then its listing, whole. Its claims,
  abstract and bibliographic data are as the office wrote them.
- **PDB 1N8Z** (CC0 1.0): trastuzumab's heavy and light chains, cut to their variable domains. The form's **CDRs typed
  in** take its Kabat CDRs as WO 2022/051390 A1 gives them: H1 DTYIH, H2 RIYPTNGYTRYADSVKG, H3 WGGDGFYAMDY, L1
  RASQDVNTAVA, L2 SASFLYS and L3 QQHYTTPPT.
- **NCBI RefSeq NM_000207.3**, bases 60 to 392: ATG to the TAG stop, which code for preproinsulin, UniProt P01308.
- **UniProtKB P01275**, human pro-glucagon (sequence version 3, entry version 242), as UniProt gave it on 15 November
  2026: residues 98 to 128, glucagon-like peptide 1(7-37), the first query of `blastp-demo.csv`. Its other two
  queries are patent proteins: `ACH25305.1`, SEQ ID NO 5223 of US 7405291-A, and `AOB58522.1`, SEQ ID NO 287 of
  US 9109239-B2.
- The database of that run: the 58 published patent protein records of Sequences' test snapshot, cut unchanged from
  GenBank release 274.0 (patent division) and from the patent protein files of EMBL-EBI, and **UniProtKB/Swiss-Prot
  Q6GZX4** (001R_FRG3G, release 2026_03), which no patent holds.

## How each file was made

- FASTA and PIR files hold each sequence's letters as the listing gives them. A header names the publication, the SEQ
  ID NO, the listing's note and the organism. Queries are named `SEQ_ID_NO_<n>`, which Sequences reads as that SEQ ID
  NO when you import a result file into a case. The database's sequences are named `<publication><kind>_<n>`, as
  Sequences names a patent sequence, so the result file's rows can join stored sequences.
- `us12788009b2-claims.txt` holds the claims as Sequences' reader reads them from the grant's XML, one after another
  with a blank line between. `[image]` stands where the claim shows a drawing of a chemical formula.
- The patterns read claim 1: "Y1 is absent or is a peptide comprising from 1 to 2 amino acids; Z is a peptide
  according to formula: AIMARPQVPRL (SEQ ID NO:1); Y2 is absent". With R1 at the start, as in claim 5's
  "[Pyr]-AIMARPQVPRL-[NH2]", that is up to two amino acids and then AIMARPQVPRL, which ends the sequence.
- `us12788009b2-blastp.tsv` is the output of a real run of BLAST+ 2.17.0 (package build of 1 July 2025, Windows x64)
  on 2026-11-15, with these commands:

  ```
  makeblastdb -in s1-patent-proteins.fasta -dbtype prot -parse_seqids -out s1
  blastp -query us12788009b2-peptides.fasta -db s1 -task blastp-short -evalue 1000 -max_target_seqs 5 -outfmt 6
  ```

  Most of its rows find the grant's own peptides, as a search of a patent's own sequences does.
- `blastp-demo.csv` is the output of a real run of BLAST+ 2.17.0 (package build of 1 July 2025, Windows x64) on
  2026-11-15 by `scripts/sequences-column-map-files.mjs`, with `BLAST_USAGE_REPORT=false` and these commands, where
  `demo-queries.fa` holds the bytes of `blastp-demo-queries.fasta`:

  ```
  makeblastdb -in library-proteins.fa -dbtype prot -parse_seqids -out demo-proteins -title "Patent proteins"
  blastp -query demo-queries.fa -db demo-proteins -outfmt "7 qaccver saccver pident length mismatch gapopen qstart qend sstart send evalue bitscore qcovhsp"
  blastp -query demo-queries.fa -db demo-proteins -outfmt "10 qaccver saccver pident length mismatch gapopen qstart qend sstart send evalue bitscore qcovhsp"
  ```

  Its header row is the `# Fields:` line of the first run, its names joined by commas, and its 14 rows are the second
  run's output as BLAST+ wrote it, CRLF line ends and all. Matched by their names, `query acc.ver` is the query,
  `subject acc.ver` the hit, `% identity` the identity, `% query coverage per hsp` the coverage and `bit score` the
  score; the other 8 columns stay with each row. On a server whose patent sets hold these accessions, 12 rows join;
  the 2 that name `Q6GZX4` do not, since no patent set holds it.

## Terms

The patent files are the US patent office's public patent data. The terms of use of its websites say that "the text
and drawings of a patent are typically not subject to copyright restrictions", subject to the limited exceptions in
37 CFR 1.71(d) and (e) and 1.84(s), and ask reusers to credit the office as the source, as this file does. PDB 1N8Z
is under CC0 1.0. The RefSeq coding sequence is the one Sequences' **Suggested searches** already search. The UniProtKB
entries behind the demo CSV, P01275 and Q6GZX4, are under CC BY 4.0, from The UniProt Consortium, and GenBank's
patent division is in the public domain.
