---
title: BLAST
description: Align a nucleotide or protein query against NCBI databases and score the hits, locally or remotely from a job, then fold the top hit with ESMFold.
tags:
  - blast
---

BLAST finds sequences similar to a query by aligning it against biological databases and scoring the matches for statistical significance. Reach for it in structure work to identify homologs of your protein — related sequences whose known structures or annotations you can borrow, or whose top hit you can fold. It is a search step that feeds the folding tools rather than producing structures itself: pair it with [ESMFold](/docs/esmfold) to fold a hit, or [AlphaFold](/docs/alphafold) and the [PDB](/docs/pdb) to look one up. See [what you can achieve](/docs/capabilities#structure) for the full set.

**BLAST (Basic Local Alignment Search Tool)** is the gold standard for biological sequence similarity searching. Developed by NCBI, BLAST compares nucleotide or protein sequences against databases to identify similar sequences and calculate statistical significance.

## Citations

> Altschul SF, Gish W, Miller W, Myers EW, Lipman DJ. Basic local alignment search tool. *Journal of Molecular Biology*. 1990 Oct 5;215(3):403-10. doi: [10.1016/S0022-2836(05)80360-2](https://doi.org/10.1016/S0022-2836(05)80360-2)

> Camacho C, Coulouris G, Avagyan V, Ma N, Papadopoulos J, Bealer K, Madden TL. BLAST+: architecture and applications. *BMC Bioinformatics*. 2009 Dec 15;10:421. doi: [10.1186/1471-2105-10-421](https://doi.org/10.1186/1471-2105-10-421)

> **NCBI BLAST**: [https://blast.ncbi.nlm.nih.gov](https://blast.ncbi.nlm.nih.gov)

## Recent updates (2024)

- **ClusteredNR**: The default database for protein BLAST searches is transitioning to NCBI ClusteredNR, built from protein sequence clusters
- **Core nucleotide (core_nt)**: An alternative to the default nt database that is more compact and efficient—less than half the size

Beyond the hosted search, these guides cover BLAST+ command lines — remote searches, preparing databases, and repeatable batch runs with clean outputs. You can ask Ivy to run them in the chat's [code workspace](/docs/sandbox), which ships BLAST+, or run them on your own machine.

## Chat Usage

### Example Prompts

- "Run BLAST on this protein sequence against SwissProt"
- "Search for homologs of human insulin in the PDB database"
- "BLAST this sequence and fold the top hit with ESMFold"

### Workflow Chaining

- **Feeds into**: ESMFold (fold hit sequences), Target Passport (identified targets), UniProt (annotation lookup)
- **Accepts from**: UniProt (sequences), ESMFold (query protein identity)

## FAQ

### Should I run BLAST locally or remotely?

  Use local BLAST for large/private databases; use remote BLAST for quick public searches without bundling databases.

### How do I prepare databases?

  Use `makeblastdb` for FASTA inputs and store databases under `inputs/db/` or a mounted S3 bucket.

### How do I keep runs reproducible?

  Pin BLAST+ version, log database names and checksums, and keep command-line arguments in the output summary.

## Read next