---
title: UniProt Search inputs and parameters
description: 'Ask Ivy in chat with an accession, a gene symbol or a free-text query, name an organism to filter by, and ask for the sequence when you need it.'
tags:
  - uniprot
  - inputs
---

You search by accession ID, gene symbol, or free-text query. An accession is the most specific — it returns the exact entry — while gene and free-text searches depend on the organism filter to disambiguate which protein you get back.

## Input methods

You search UniProt by asking Ivy in chat. There is no form, job, or command for it. (For scripted access, UniProt also has a read-only [REST lookup](/docs/api/rest-lookup).)

## Chat input (natural language)

Type a protein query in plain English. The orchestrator detects UniProt intent and determines the appropriate search strategy.

**Examples:**
```
Search UniProt for human EGFR
Get the sequence for P04637 from UniProt
What does the BRCA1 protein do?
Find human kinases involved in apoptosis
```

The orchestrator extracts protein names, gene symbols, accession IDs, and organism information from your prompt and maps them to the appropriate UniProt search.

## What you can ask for

### Full protein record

The fuller of the two UniProt searches. Give at least one of an accession, a gene symbol, or a free-text description.

| You can set | Default | Limits |
|-------------|---------|--------|
| Accession | None | A UniProt accession ID (e.g., P04637, Q9Y6K9) — the most specific lookup |
| Gene symbol | None | e.g., TP53, EGFR |
| Free text | None | Protein name, function, or disease |
| Organism | All organisms | An organism's scientific name; applied only when you name one |
| Sequence | Not included | Ask for the sequence and the full amino acid sequence comes back |
| Features | Included | Domains, sites, and PTMs |
| Number of results | 5 | 1-25 |

### Quick search

A lighter keyword search across protein name, gene, accession, or keyword, returning up to 100 entries (25 by default).

## Input strategies

### Accession lookup (fastest, most specific)

When you have a UniProt accession ID, give it in your question for a direct entry lookup. This bypasses the search index and returns the exact protein record.

**Accession format**: UniProt accessions follow specific patterns:
- **SwissProt**: 6 characters (e.g., P00533, Q9Y261)
- **TrEMBL**: 6 or 10 characters (e.g., A0A0A0MRZ7)

```
Get the UniProt entry for P00533 with its sequence and features
```

### Gene-based search

Search by gene symbol when you know the gene but not the UniProt accession. Naming the organism (the search covers all organisms unless you do) helps disambiguate genes present in multiple species.

```
Find human EGFR in UniProt and include the sequence
```

### Free-text search

Describe the protein in your own words for broad searches by protein name, function, disease, or any keyword. This searches across all UniProt fields.

```
Search UniProt for human tyrosine kinase receptors involved in cell growth, 10 results
```

### Cross-species search

Name the organism to get a specific species' entry:

```
Get mouse (Mus musculus) TP53 from UniProt with its sequence
```

## Organism names

Use the scientific name for the organism filter. Common examples:

| Common name | Scientific name |
|-------------|----------------|
| Human | `Homo sapiens` |
| Mouse | `Mus musculus` |
| Rat | `Rattus norvegicus` |
| Zebrafish | `Danio rerio` |
| Fruit fly | `Drosophila melanogaster` |
| Yeast | `Saccharomyces cerevisiae` |
| E. coli | `Escherichia coli` |

## Input validation

### Accession validation

- **Format**: Must match UniProt accession patterns (6 or 10 alphanumeric characters)
- **Case**: Case-insensitive (p00533 is accepted and converted to P00533)
- **Invalid accessions**: Return an error with suggestion to verify on UniProt website

### Gene symbol validation

- **Format**: Free text, typically uppercase gene symbols
- **Ambiguous symbols**: Some gene symbols map to multiple proteins. The organism filter helps disambiguate.
- **Aliases**: UniProt indexes gene name synonyms, so common aliases usually work

### Query validation

- **Minimum**: At least one of accession, gene, or query must be provided
- **Empty input**: Returns an error with guidance
- **Special characters**: Most characters are safe; UniProt's search supports quoted phrases

### Limit validation

- **Minimum**: 1
- **Maximum**: 25 for a full protein record search, 100 for a quick search
- **Default**: 5 for a full protein record search, 25 for a quick search
- **Values outside range**: Clamped to nearest boundary

## Example prompts

```
Retrieve UniProt P04637 with its sequence and features
Search UniProt for human EGFR-family receptor tyrosine kinases, with features, 10 results
Find the mouse ortholog of human BRCA1 in UniProt and give me its sequence
Search UniProt for human proteins associated with Alzheimer disease amyloid, 10 results
```

## Read next