search_records
Read-only
Search the sequence archives by free text (a gene name, a protein, an organism, a phrase) across NCBI Nucleotide and Protein, ENA (plain records, WGS and TSA/TLS assembly sets, and annotated coding/non-coding features), UniProtKB and the iGEM Registry at once, and hand back the matching accessions. This is the half of Fetch from database that answers "which record do I want?"; fetch_record answers "give me that one". A source that is down does not fail the search: its error comes back beside the hits the others returned.
Input Schema
TableJSON Schema
| Name | Required | Description | Default |
|---|---|---|---|
| limit | No | how many hits to ask each source for. The panel pages at this size, and every source is asked for the same number. | |
| query | Yes | the words to search for. Archive query syntax is accepted where the archive accepts it, but plain words are the normal case: "beta-lactamase Escherichia coli", "GFP", "pUC19". | |
| _files | No | The project: record references shaped like _subject. Their records become library documents a tool names in sequences/parts/document; a trace or alignment is found by file name. | |
| sources | No | which archives to ask. Each is searched independently and the hits are interleaved, so asking more costs time rather than accuracy. Defaults to the nucleotide pair the panel opens on. | |
| _subject | No | The record to act on: exactly one of sequence, accession (+source), sample (an id from list_samples) or file ({name, text} or {name, base64}); record_index picks one record of several. |
Output Schema
TableJSON Schema
| Name | Required | Description | Default |
|---|---|---|---|
| hits | Yes | the full result rows: accession, title, source, organism, length, molecule kind and the archive’s own reported fields | |
| source | Yes | which archive the top hit came from, ready for the same node's $source where the two vocabularies agree | |
| hitCount | Yes | how many records came back across every source asked | |
| accession | Yes | the top hit's accession, ready for a fetch_record node's $accession. This is the seam between searching and fetching | |
| accessions | Yes | every hit’s accession, newline-separated, in the order they are ranked | |
| unavailable | Yes | the sources that could not answer, each with the reason: empty when every source answered. A search is not a failure because one archive is down, and hiding that would make a short list read as a small literature |