UCSC Genome Browser

Overview

The UCSC Genome Browser REST API at https://api.genome.ucsc.edu/ provides programmatic access to genome sequences, annotation tracks, and hub data for 100+ assemblies including hg38, mm39, and dm6. The API is free, requires no authentication, and returns JSON. Use it with the requests library to fetch DNA sequences for genomic regions, retrieve track data (genes, repeats, conservation), list available tracks, and query chromosome sizes for genome-scale coordinate arithmetic.

When to Use

• Fetching the reference DNA sequence for any genomic region (e.g., promoter, exon, CRISPR target) across human, mouse, or other assemblies
• Retrieving RefSeq or GENCODE gene structure (exon coordinates, CDS boundaries, strand) for a locus of interest
• Looking up PhyloP or PhastCons conservation scores to assess evolutionary constraint at a variant site
• Listing and querying any of UCSC's 1000+ annotation tracks (repeats, regulatory elements, conservation) for a region
• Getting chromosome sizes for a genome assembly to set up bedtools, pysam, or coverage pipelines
• Accessing public UCSC track hubs (e.g., ENCODE, Roadmap Epigenomics) without downloading data locally
• Use ensembl-database instead when you need Ensembl stable IDs, VEP variant annotation, or cross-species comparative genomics via the Ensembl REST API
• For bulk local queries across millions of regions, use bedtools-genomic-intervals with pre-downloaded UCSC annotation files

Prerequisites

• Python packages: requests, matplotlib (for visualization)
• Data requirements: genomic coordinates (chrom, start, end in 0-based half-open BED format), genome assembly name (e.g., hg38, mm39)
• Environment: internet connection; no authentication required
• Rate limits: no official published limit; add 0.5s delays for batch requests (>100 queries)

pip install requests matplotlib

Quick Start

import requests

BASE = "https://api.genome.ucsc.edu"

def get_sequence(genome, chrom, start, end):
    """Fetch DNA sequence for a genomic region (0-based, half-open)."""
    r = requests.get(f"{BASE}/getData/sequence",
                     params={"genome": genome, "chrom": chrom,
                             "start": start, "end": end})
    r.raise_for_status()
    return r.json()["dna"]

# Fetch 1 kb around the BRCA1 TSS on hg38
seq = get_sequence("hg38", "chr17", 43044294, 43045294)
print(f"Length: {len(seq)} bp")
print(f"Sequence: {seq[:60]}...")
# Length: 1000 bp
# Sequence: ATGATTGGTGGTTACATGCACAGTTGCTCTGGGAAGTTTCTTCTTCAGTTGAGAAAAGGT...

Core API

Query 1: Sequence Retrieval

Fetch the reference DNA sequence for any genomic region using the getData/sequence endpoint. Coordinates are 0-based, half-open (BED format).

import requests

BASE = "https://api.genome.ucsc.edu"

def get_sequence(genome, chrom, start, end):
    """Return DNA sequence string for the given region."""
    r = requests.get(f"{BASE}/getData/sequence",
                     params={"genome": genome, "chrom": chrom,
                             "start": start, "end": end})
    r.raise_for_status()
    data = r.json()
    return data["dna"]

# TP53 exon 4 region (hg38)
seq = get_sequence("hg38", "chr17", 7676520, 7676620)
print(f"Region: chr17:7,676,520-7,676,620 ({len(seq)} bp)")
print(f"Sequence: {seq}")

# Reverse-complement for minus-strand genes
def revcomp(seq):
    comp = str.maketrans("ACGTacgt", "TGCAtgca")
    return seq.translate(comp)[::-1]

# BRCA2 on minus strand (hg38)
seq_fwd = get_sequence("hg38", "chr13", 32315086, 32315186)
seq_rc  = revcomp(seq_fwd)
print(f"Forward: {seq_fwd[:30]}...")
print(f"RevComp: {seq_rc[:30]}...")

Query 2: Track Data Query

Retrieve annotation data (BED records) from any UCSC track for a genomic region.

import requests

BASE = "https://api.genome.ucsc.edu"

def get_track_data(genome, track, chrom, start, end):
    """Fetch annotation records from a UCSC track for a region."""
    r = requests.get(f"{BASE}/getData/track",
                     params={"genome": genome, "track": track,
                             "chrom": chrom, "start": start, "end": end})
    r.raise_for_status()
    data = r.json()
    # Track data is under the key matching the track name
    return data.get(track, data.get("data", []))

# Fetch RepeatMasker annotations in the MYC locus (hg38)
repeats = get_track_data("hg38", "rmsk", "chr8", 127_735_434, 127_742_951)
print(f"Repeat elements in MYC locus: {len(repeats)}")
for r in repeats[:3]:
    print(f"  {r.get('repName', r.get('name'))} | {r['chromStart']}-{r['chromEnd']}")

# Fetch CpG islands near a promoter
cpg_islands = get_track_data("hg38", "cpgIslandExt", "chr17", 43_044_000, 43_050_000)
print(f"CpG islands found: {len(cpg_islands)}")
for island in cpg_islands:
    print(f"  {island['name']}: {island['chromStart']}-{island['chromEnd']}, "
          f"obsExp={island.get('obsExp', 'n/a')}")

Query 3: Track List

List all available annotation tracks for a genome assembly to discover what data is available.

import requests

BASE = "https://api.genome.ucsc.edu"

def list_tracks(genome):
    """Return a dict of {track_name: track_metadata} for a genome assembly."""
    r = requests.get(f"{BASE}/list/tracks", params={"genome": genome})
    r.raise_for_status()
    return r.json().get("tracks", {})

tracks = list_tracks("hg38")
print(f"Total tracks in hg38: {len(tracks)}")

# Find conservation-related tracks
conserv = {k: v for k, v in tracks.items() if "conserv" in k.lower() or "phylop" in k.lower()}
for name, meta in list(conserv.items())[:5]:
    print(f"  {name}: {meta.get('shortLabel', '')}")

Query 4: Chromosome Sizes

Get the length of every chromosome (or scaffold) for a genome assembly.

import requests

BASE = "https://api.genome.ucsc.edu"

def get_chrom_sizes(genome):
    """Return {chrom: size_in_bp} for a genome assembly."""
    r = requests.get(f"{BASE}/list/chromosomes", params={"genome": genome})
    r.raise_for_status()
    return r.json().get("chromosomeSizes", {})

sizes = get_chrom_sizes("hg38")
print(f"hg38 chromosome count: {len(sizes)}")

# Show canonical autosomes + sex chromosomes
canonical = {c: sizes[c] for c in sorted(sizes) if c in
             [f"chr{i}" for i in range(1, 23)] + ["chrX", "chrY", "chrM"]}
for chrom, length in sorted(canonical.items(),
                             key=lambda x: int(x[0].replace("chr", "").replace("X", "23").replace("Y", "24").replace("M", "25"))):
    print(f"  {chrom}: {length:,} bp")

Query 5: Gene Annotation

Query RefSeq gene models (exon coordinates, CDS, strand) for a genomic region.

import requests

BASE = "https://api.genome.ucsc.edu"

def get_refgene(genome, chrom, start, end):
    """Retrieve RefSeq gene annotations for a region."""
    r = requests.get(f"{BASE}/getData/track",
                     params={"genome": genome, "track": "refGene",
                             "chrom": chrom, "start": start, "end": end})
    r.raise_for_status()
    return r.json().get("refGene", [])

# Query EGFR gene region (hg38)
genes = get_refgene("hg38", "chr7", 55_019_017, 55_211_628)
for g in genes:
    exon_count = g.get("exonCount", 0)
    print(f"  {g['name2']} ({g['name']}) | {g['strand']} | "
          f"tx: {g['txStart']}-{g['txEnd']} | exons: {exon_count}")

# Parse exon intervals from a refGene record
def parse_exons(gene_record):
    """Return list of (exon_start, exon_end) from a refGene record."""
    starts = [int(s) for s in gene_record["exonStarts"].strip(",").split(",") if s]
    ends   = [int(e) for e in gene_record["exonEnds"].strip(",").split(",") if e]
    return list(zip(starts, ends))

genes = get_refgene("hg38", "chr7", 55_019_017, 55_211_628)
if genes:
    g = genes[0]
    exons = parse_exons(g)
    print(f"{g['name2']}: {len(exons)} exons")
    for i, (s, e) in enumera