NCBI Genome Assemblies Scraper

Pull structured genome assembly reports straight from the NCBI Datasets v2 API — assembly quality stats (N50, contig & scaffold counts, total length, GC%), organism/taxonomy, BioProject, submitter and annotation gene counts — by organism/taxon or by assembly accession. No API key, no login, no scraping HTML.

Runs on ~$0.50 per 1,000 assemblies of compute — the data itself is free and public.

What you get

For every assembly, one clean JSON record:

Field	Example
accession / current_accession	GCF_000001405.40
organism, tax_id, common_name	Homo sapiens, 9606, human
assembly_name, assembly_level	GRCh38.p14, Chromosome
assembly_type, refseq_category	haploid-with-alt-loci, reference genome
bioproject, submitter, submission_date	PRJNA31257, Genome Reference Consortium, 2022-02-03
total_length_bp, gc_percent, chromosome_count	3099441038, 41.0, 24
contig_count, contig_n50	contiguity metrics
scaffold_count, scaffold_n50	scaffolding metrics
gene_total, protein_coding_genes	59792, 20076
annotation_provider, annotation_release_date	NCBI RefSeq, 2025-08-01
url, source, scraped_at	canonical NCBI Datasets page + metadata

All numeric fields are real numbers (not strings). Every field is nullable.

Input

Field	Type	Default	Notes
mode	select	taxon	taxon (all assemblies for organisms) or accessions (explicit accessions)
taxa	array	["human","Escherichia coli","Arabidopsis thaliana"]	organism names, common names or Tax IDs
accessions	array	["GCF_000001405.40"]	GenBank/RefSeq assembly accessions
assemblyLevel	select	any	any / complete / chromosome / scaffold / contig (taxon mode)
referenceOnly	boolean	false	only RefSeq reference/representative genomes (taxon mode)
maxItems	integer	200	cap across all inputs (max 5000)

Example

{
  "mode": "taxon",
  "taxa": ["Escherichia coli"],
  "assemblyLevel": "complete",
  "referenceOnly": false,
  "maxItems": 500
}

Use cases

• Comparative genomics — pull every assembly for a clade and compare length, GC% and gene counts side by side.
• Assembly QC — screen contig/scaffold N50 and contig counts to flag fragmented or low-quality assemblies.
• Bioinformatics pipelines — resolve accessions to metadata (organism, BioProject, annotation) as a pipeline lookup step.
• Reference selection — use referenceOnly to grab the canonical reference/representative genome per organism.

How is this different from ncbi-eutils-scraper?

Both hit NCBI, but they are different products for different data:

• ncbi-eutils-scraper uses the classic E-utilities (esearch/efetch/esummary). It is built for literature and gene search — PubMed articles, Gene records, and other text/entity databases.
• This actor (NCBI Genome Assemblies Scraper) uses the modern NCBI Datasets v2 API. It is built for genome assemblies and their statistics — N50, contig/scaffold counts, total length, GC%, BioProject and annotation gene counts, returned as flat typed records.

Use the E-utils actor when you want papers or gene entries; use this one when you want assembly-level genome data and quality metrics.

Related actors

• ncbi-eutils-scraper — PubMed / Gene literature and entity search
• ensembl-scraper — Ensembl genes, transcripts and cross-references
• uniprot-scraper — UniProt protein entries
• interpro-scraper — InterPro protein domains and families

Notes

Data comes from the free public NCBI Datasets v2 API (https://api.ncbi.nlm.nih.gov/datasets/v2alpha) with no API key required. Please respect NCBI's usage policy. This actor is not affiliated with or endorsed by NCBI/NLM/NIH.