regulomedb-database

name	regulomedb-database
description	Query RegulomeDB v2 REST API to score variants for regulatory function and retrieve overlapping evidence (TF binding, histone marks, DNase peaks, eQTLs, motifs). Score single rsID/position, batch lists, region searches, and full annotations. Scores range 1a (strongest: eQTL+TF+DNase+motif) to 7 (none). Use for GWAS hit prioritization, regulatory variant annotation, cis-regulatory discovery. Use clinvar-database for pathogenicity; gwas-database for trait associations.
license	CC-BY-4.0

RegulomeDB Database

Overview

RegulomeDB integrates large-scale functional genomics data (ENCODE, Roadmap Epigenomics) to score genetic variants for regulatory potential. Each variant receives a score from 1a (highest regulatory confidence: eQTL + TF binding + DNase accessibility + motif + chromatin) to 7 (no known regulatory function). The v2 REST API supports single-variant queries, batch scoring, and region-based searches. Access is free and requires no authentication.

When to Use

• Prioritizing GWAS hits for regulatory follow-up — identify which SNPs land in active regulatory elements
• Annotating a VCF or variant list with regulatory scores to filter to functionally relevant variants
• Identifying which transcription factors bind near a variant of interest
• Checking whether a non-coding variant overlaps an eQTL and active chromatin simultaneously
• Retrieving all high-confidence regulatory variants in a genomic region for cis-regulatory analysis
• Use clinvar-database instead when you need clinical pathogenicity classifications (ClinSig); RegulomeDB is for regulatory function, not germline disease association
• Use gwas-database instead when you want published GWAS associations with traits; RegulomeDB scores regulatory evidence regardless of trait association

Prerequisites

• Python packages: requests, pandas, matplotlib
• Data requirements: rsIDs (e.g., rs4946036), genomic positions (chr1:1000000), or region coordinates (chr1:1000000-2000000)
• Genome build: GRCh38 (default) or GRCh37; specify in all requests
• Rate limits: No published rate limits; use time.sleep(0.3) between requests in batch workflows

pip install requests pandas matplotlib

Quick Start

import requests

def regulome_score(variant, genome="GRCh38"):
    """Score a single variant (rsID or chr:pos) for regulatory function."""
    r = requests.post(
        "https://regulomedb.org/regulome-search/",
        json={"variants": [variant], "genome": genome, "limit": 1},
        headers={"Content-Type": "application/json"}
    )
    r.raise_for_status()
    data = r.json()
    hits = data.get("variants", [])
    if not hits:
        return None
    v = hits[0]
    return {"rsid": variant, "score": v.get("regulomedb_score"), "chrom": v.get("chrom"), "pos": v.get("start")}

result = regulome_score("rs4946036")
print(f"Variant: {result['rsid']}  Score: {result['score']}  Position: {result['chrom']}:{result['pos']}")
# Variant: rs4946036  Score: 1b  Position: chr5:1295228

Core API

Query 1: Variant Scoring — Single rsID or Genomic Position

Post a single variant to the main search endpoint and retrieve its regulatory score.

import requests

BASE = "https://regulomedb.org"

def score_variant(variant, genome="GRCh38"):
    """
    Score a single variant for regulatory potential.
    variant: rsID (e.g. 'rs4946036') or chr:pos (e.g. 'chr5:1295228')
    Returns: dict with score, coordinates, and hit count.
    """
    r = requests.post(
        f"{BASE}/regulome-search/",
        json={"variants": [variant], "genome": genome, "limit": 10},
        headers={"Content-Type": "application/json"},
        timeout=30
    )
    r.raise_for_status()
    data = r.json()
    hits = data.get("variants", [])
    if not hits:
        print(f"{variant}: no RegulomeDB annotation found")
        return None
    v = hits[0]
    print(f"Variant  : {v.get('rsids', [variant])}")
    print(f"Score    : {v.get('regulomedb_score')}")
    print(f"Location : {v.get('chrom')}:{v.get('start')}-{v.get('end')}")
    print(f"Evidence : {v.get('num_peaks', 0)} overlapping peaks")
    return v

result = score_variant()

# Query by genomic coordinate instead of rsID
result = score_variant("chr17:43092919", genome="GRCh38")
# Covers TP53 region

Query 2: Batch Variant Scoring — Multiple Variants

Score a list of variants in a single POST request; handle large lists in batches of 100.

import requests, time, pandas as pd

BASE = "https://regulomedb.org"

def batch_score_variants(variants, genome="GRCh38", batch_size=100):
    """
    Score multiple variants and return a DataFrame of results.
    variants: list of rsIDs or chr:pos strings
    Returns: pd.DataFrame with columns [variant, score, chrom, pos, num_peaks]
    """
    records = []
    for i in range(0, len(variants), batch_size):
        batch = variants[i:i + batch_size]
        r = requests.post(
            f"{BASE}/regulome-search/",
            json={"variants": batch, "genome": genome, "limit": batch_size},
            headers={"Content-Type": "application/json"},
            timeout=60
        )
        r.raise_for_status()
        data = r.json()
        for v in data.get("variants", []):
            records.append({
                "query": batch[len(records) % batch_size] if len(records) < len(batch) else "",
                "rsids": "; ".join(v.get("rsids", [])),
                "score": v.get("regulomedb_score"),
                "chrom": v.get("chrom"),
                "pos": v.get("start"),
                "num_peaks": v.get("num_peaks", ),
            })
         i + batch_size < (variants):
            time.sleep()

    df = pd.DataFrame(records)
    ()
    (df[].value_counts().sort_index())
     df

gwas_hits = [
    , , , ,
    , , ,  
]
df = batch_score_variants(gwas_hits)
df.to_csv(, index=)
(df[[, , , ]].head())

Query 3: Region Search — Find Regulatory Variants in a Genomic Window

Retrieve all annotated variants within a chromosomal region using the GET endpoint.

import requests, pandas as pd

BASE = "https://regulomedb.org"

def search_region(chrom, start, end, genome="GRCh38", limit=200):
    """
    Find all RegulomeDB-annotated variants in a genomic region.
    Returns: pd.DataFrame of variants with scores.
    """
    params = {
        "regions": f"{chrom}:{start}-{end}",
        "genome": genome,
        "limit": limit,
    }
    r = requests.get(f"{BASE}/regulome-search/", params=params, timeout=60)
    r.raise_for_status()
    data = r.json()
    variants = data.get("variants", [])
    print(f"Found {len(variants)} annotated variants in {chrom}:{start}-{end}")

    records = []
    for v in variants:
        records.append({
            "rsids": "; ".join(v.get("rsids", [])),
            "score": v.get("regulomedb_score"),
            "chrom": v.get("chrom"),
            "start": v.get("start"),
            "end": v.get("end"),
            "num_peaks": v.get("num_peaks", 0),
        })
    return pd.DataFrame(records)

# Search the BRCA1 promoter region (GRCh38)
df_region = search_region(, , )

high_conf = df_region[df_region[].isin([, , , ])]
()
(df_region.sort_values().head().to_string(index=))

Query 4: Annotation Details — Full Regulatory Evidence for a Variant

Retrieve the complete evidence set for a variant: TF names, histone marks, eQTL status, motif hits.

import requests

BASE = "https://regulomedb.org"

def get_annotation_details(variant, genome="GRCh38"):
    """
    Fetch full regulatory annotation details for a single variant.
    Returns raw JSON response with peak-level evidence.
    """
    r = requests.post(
        f"{BASE}/regulome-search/",
        json={"variants": [variant], "genome": genome, "limit": 1},
        headers={"Content-Type": "application/json"},
        timeout=30
    )
    r.raise_for_status()
    data = r.json()
    hits = data.get("variants", [])
    if not hits:
        print(f"No annotation found for {variant}")
        return None

    v = hits[0]
    print(f"=== {variant} | Score: {v.get('regulomedb_score')} ===")

    # Evidence breakdown
    peaks = v.get("peaks", [])
    tf_peaks = [p for p in peaks if p.get("assay_type") == "TF ChIP-seq"]
    dnase_peaks = [p for p in peaks if p.get("assay_type") == "DNase-seq"]
    histone_peaks = [p for p in peaks if "histone" in p.get(, ).lower()]

    ()
    ()
    ()

    
    tfs = ((p.get(, {}).get(, )  p  tf_peaks))
    ()

    eqtls = v.get(, [])
    ()
     eq  eqtls[:]:
        ()

     v

detail = get_annotation_details()

Query 5: Summary Statistics — Score Distribution in a Region

Get counts of variants by regulatory score category using the summary endpoint.

import requests

BASE = "https://regulomedb.org"

def get_region_summary(chrom, start, end, genome="GRCh38"):
    """
    Retrieve summary statistics: count of variants per score category in a region.
    Returns: dict mapping score → count.
    """
    r = requests.post(
        f"{BASE}/regulome-summary/",
        json={"regions": [f"{chrom}:{start}-{end}"], "genome": genome},
        headers={"Content-Type": "application/json"},
        timeout=30
    )
    r.raise_for_status()
    data = r.json()
    summary = data.get("summary", {})
    print(f"Regulatory score summary for {chrom}:{start}-{end}")
    score_order = ["1a", "1b", "2a", "2b", "2c", "3a", "3b", "4", "5", "6", "7"]
    for score in score_order:
        count = summary.get(score, 0)
        bar = "#" * min(count, 40)
        print(f"  Score {score:3s}: {count:5d}  {bar}")
    total = sum(summary.values())
    print(f"  Total annotated: ")
     summary


summary = get_region_summary(, , )

Query 6: Dataset Information — Available Regulatory Datasets

List the regulatory datasets (cell types, assay types) included in RegulomeDB.

import requests, pandas as pd

BASE = "https://regulomedb.org"

def list_datasets(assay_type=None):
    """
    List available regulatory datasets (cell types and assay types).
    assay_type: optional filter, e.g. 'TF ChIP-seq', 'DNase-seq', 'Histone ChIP-seq'
    Returns: pd.DataFrame of datasets.
    """
    params = {"format": "json"}
    if assay_type:
        params["assay_type"] = assay_type

    r = requests.get(f"{BASE}/regulome-datasets/", params=params, timeout=30)
    r.raise_for_status()
    data = r.json()
    datasets = data.get("datasets", [])
    print(f"Total datasets: {len(datasets)}")

    records = []
    for ds in datasets[:20]:  # show first 20
        records.append({
            "dataset_id": ds.get("@id", "").split("/")[-2],
            "assay_type": ds.get("assay_type"),
            "biosample": ds.get("biosample_term_name"),
            "target": ds.get("target", {}).get("label", ""),
        })
    df = pd.DataFrame(records)
    print(df.to_string(index=False))
    return df

df_datasets = list_datasets(assay_type="TF ChIP-seq")
print(f"\nUnique cell types: {df_datasets[].nunique()}")

Key Concepts

RegulomeDB Scoring Schema

RegulomeDB scores encode the type and strength of regulatory evidence overlapping a variant:

Score	Evidence	Regulatory Confidence
1a	eQTL + TF binding + DNase + motif + chromatin	Highest
1b	TF binding + DNase + motif + chromatin (no eQTL)	Very high
2a	TF binding + DNase + motif	High
2b	TF binding + DNase (no motif)	High
2c	TF binding only (with DNase)	Moderate-high
3a	DNase + motif	Moderate
3b	Motif only	Moderate
4	Single TF binding evidence	Low-moderate
5	Chromatin accessibility only	Low
6	Other regulatory evidence	Minimal
7	No known regulatory function	None

Scores 1a–2b are the most actionable for prioritizing GWAS hits or regulatory variant interpretation. Score 7 means the variant has not been overlapped by any ENCODE/Roadmap regulatory dataset.

Variant Input Formats

RegulomeDB accepts three input formats for variants:

# Format 1: rsID (resolved to GRCh38 coordinates internally)
variants_rsid = ["rs4946036", "rs7903146", "rs12345678"]

# Format 2: chr:pos (single nucleotide position)
variants_pos = ["chr5:1295228", "chr10:112998251", "chr3:185511687"]

# Format 3: chr:start-end (short range, typically SNP-sized)
variants_range = ["chr17:43092919-43092920", "chr7:117548628-117548629"]

# Mix all formats in one request
mixed = variants_rsid + variants_pos

Common Workflows

Workflow 1: GWAS Hit Prioritization

Goal: Take a list of GWAS lead SNPs and identify which have strong regulatory evidence for functional follow-up.

import requests, time, pandas as pd
import matplotlib.pyplot as plt

BASE = "https://regulomedb.org"

# GWAS lead SNPs from a type 2 diabetes GWAS
gwas_snps = [
    "rs7903146",   # TCF7L2 locus
    "rs10811661",  # CDKN2A/B locus
    "rs1801282",   # PPARG locus
    "rs4946036",   # PCSK1 locus
    "rs2268177",   # HNF4A locus
    "rs11605924",  # CRY2 locus
    "rs10830963",  # MTNR1B locus
    "rs1111875",   # HHEX locus
]

records = []
for snp in gwas_snps:
    r = requests.post(
        f"{BASE}/regulome-search/",
        json={"variants": [snp], "genome": "GRCh38", "limit": 1},
        headers={"Content-Type": "application/json"},
        timeout=30
    )
    r.raise_for_status()
    data = r.json()
    hits = data.get("variants", [])
    if hits:
        v = hits[0]
        peaks = v.get("peaks", [])
        tfs = [p.get("target", {}).get("label", "") for p in peaks if p.get("assay_type") == "TF ChIP-seq"]
        records.append({
            "snp": snp,
            : v.get(),
            : v.get(, ),
            : ((tfs)),
            : (v.get(, [])) > ,
        })
    :
        records.append({: snp, : , : , : , : })
    time.sleep()

df = pd.DataFrame(records)
df[] = df[]..replace(, )..replace(, )..replace(, ).astype()
df_sorted = df.sort_values()

()
(df_sorted[[, , , , ]].to_string(index=))
()
df_sorted.to_csv(, index=)


fig, ax = plt.subplots(figsize=(, ))
score_counts = df[].value_counts().sort_index()
ax.bar(score_counts.index, score_counts.values, color=, edgecolor=)
ax.set_xlabel()
ax.set_ylabel()
ax.set_title()
plt.tight_layout()
plt.savefig(, dpi=, bbox_inches=)
()

Workflow 2: TF Binding Heatmap for a Variant Set

Goal: Retrieve annotation details for multiple variants and visualize which TFs bind near each variant.

import requests, time, pandas as pd
import matplotlib.pyplot as plt
import numpy as np

BASE = "https://regulomedb.org"

variants = ["rs4946036", "rs7903146", "rs10811661", "rs1801282", "rs2268177"]

all_tfs = set()
variant_tfs = {}

for snp in variants:
    r = requests.post(
        f"{BASE}/regulome-search/",
        json={"variants": [snp], "genome": "GRCh38", "limit": 1},
        headers={"Content-Type": "application/json"},
        timeout=30
    )
    r.raise_for_status()
    data = r.json()
    hits = data.get("variants", [])
    if not hits:
        variant_tfs[snp] = {}
        continue
    v = hits[0]
    peaks = v.get("peaks", [])
    tfs = [p.get("target", {}).get("label", "Unknown")
           for p in peaks if p.get("assay_type") == "TF ChIP-seq"]
    tf_counts = {}
    for tf in tfs:
        tf_counts[tf] = tf_counts.get(tf, 0) + 1
    variant_tfs[snp] = tf_counts
    all_tfs.update(tf_counts.keys())
    time.sleep(0.3)

# Build heatmap matrix
tf_list = sorted(all_tfs - {})[:]  
matrix = np.zeros(((variants), (tf_list)))
 i, snp  (variants):
     j, tf  (tf_list):
        matrix[i, j] = variant_tfs[snp].get(tf, )

fig, ax = plt.subplots(figsize=((, (tf_list) * ), (variants) *  + ))
im = ax.imshow(matrix, cmap=, aspect=)
ax.set_xticks(((tf_list)))
ax.set_xticklabels(tf_list, rotation=, fontsize=)
ax.set_yticks(((variants)))
ax.set_yticklabels(variants, fontsize=)
ax.set_title()
plt.colorbar(im, ax=ax, label=)
plt.tight_layout()
plt.savefig(, dpi=, bbox_inches=)
()

Key Parameters

Parameter	Endpoint	Default	Range / Options	Effect
variants	POST /regulome-search/	required	list of rsIDs or chr:pos	Variants to score
genome	All	"GRCh38"	"GRCh38", "GRCh37"	Reference genome assembly
limit	POST /regulome-search/	10	1–1000	Max variants returned per request
regions	GET /regulome-search/	required	"chrN:start-end"	Genomic region for region search
assay_type	GET /regulome-datasets/	—	"TF ChIP-seq", "DNase-seq", "Histone ChIP-seq"	Filter datasets by assay
format	GET endpoints	"json"	"json", "tsv"	Response format

Best Practices

1. Interpret scores in context: Scores 1a–2b indicate multi-evidence regulatory annotation, but do not guarantee causality. A score of 7 means absence of evidence in RegulomeDB's curated datasets, not evidence of absence of regulation.

2. Use region search for locus-level analysis: When analyzing a GWAS locus, use GET /regulome-search/?regions= to retrieve all annotated variants in the LD block, then filter by score and linkage disequilibrium data.

3. Add time.sleep(0.3) in loops: RegulomeDB has no published rate limit, but polite usage prevents server load issues that produce connection timeouts in long batch runs.

4. Cross-reference eQTL data: Variants with score 1a have RegulomeDB-integrated eQTL evidence. Always verify eQTL tissue specificity (GTEx tissue list in the eqtls field) before drawing biological conclusions.

5. Prefer rsIDs over coordinates for reproducibility: When sharing results, store rsIDs alongside coordinates — RegulomeDB's internal coordinate system is GRCh38, and position-based queries are build-specific.

Common Recipes

Recipe: Check Score for a Single rsID

When to use: Quick lookup for one variant before running a full batch analysis.

import requests

def quick_score(rsid, genome="GRCh38"):
    r = requests.post(
        "https://regulomedb.org/regulome-search/",
        json={"variants": [rsid], "genome": genome, "limit": 1},
        headers={"Content-Type": "application/json"},
        timeout=20
    )
    r.raise_for_status()
    hits = r.json().get("variants", [])
    score = hits[0].get("regulomedb_score") if hits else "not found"
    print(f"{rsid}: RegulomeDB score = {score}")
    return score

quick_score("rs4946036")   # Expected: 1b (strong regulatory evidence)
quick_score("rs12345678")  # Expected: 5 or 7 (weak/no evidence)

Recipe: Filter Variants to High-Confidence Regulatory Set

When to use: Downstream prioritization — keep only variants with strong regulatory evidence (scores 1a–2b).

import pandas as pd

df = pd.read_csv("gwas_regulatory_priority.csv")

REGULATORY_SCORES = {"1a", "1b", "2a", "2b"}
high_conf = df[df["score"].isin(REGULATORY_SCORES)].copy()
print(f"Total variants   : {len(df)}")
print(f"High-confidence  : {len(high_conf)}  ({100 * len(high_conf) / len(df):.1f}%)")
print(high_conf[["snp", "score", "num_tfs", "has_eqtl"]].to_string(index=False))
high_conf.to_csv("high_confidence_regulatory_variants.csv", index=False)

Recipe: Identify eQTL-Linked Regulatory Variants

When to use: Integrate RegulomeDB eQTL evidence to find variants that affect gene expression through regulatory elements.

import requests, time

BASE = "https://regulomedb.org"

def get_eqtl_variants(variant_list, genome="GRCh38"):
    """Return variants that have both high regulatory score and eQTL evidence."""
    eqtl_variants = []
    for snp in variant_list:
        r = requests.post(
            f"{BASE}/regulome-search/",
            json={"variants": [snp], "genome": genome, "limit": 1},
            headers={"Content-Type": "application/json"},
            timeout=30
        )
        r.raise_for_status()
        hits = r.json().get("variants", [])
        if not hits:
            time.sleep(0.3)
            continue
        v = hits[0]
        eqtls = v.get("eqtls", [])
        score = v.get("regulomedb_score", "7")
        if eqtls and score in {"1a", "1b", "2a", "2b"}:
            for eq in eqtls:
                eqtl_variants.append({
                    "snp": snp,
                    "score": score,
                    "gene": eq.get("gene_name"),
                    "tissue": eq.get("tissue"),
                    "slope": eq.get("slope"),
                })
        time.sleep()
     eqtl_variants

snps = [, , ]
results = get_eqtl_variants(snps)
 r  results:
    ()

Troubleshooting

Problem	Cause	Solution
HTTP 400 on POST request	Malformed JSON body or missing Content-Type header	Add headers={"Content-Type": "application/json"} and use json= kwarg, not data=
Empty variants list in response	Variant not in RegulomeDB index	Try alternate input format (rsID → chr:pos, or vice versa); some rare variants are absent
ConnectionError or timeout	Server timeout on large batch	Reduce batch size to 50; add time.sleep(0.5) between batches
Score is "7" for all variants	Using GRCh37 coordinates with genome="GRCh38"	Specify "genome": "GRCh37" or liftover coordinates to GRCh38 before querying
peaks field is empty but score is not "7"	Score assigned from summary-level evidence, not peak-level	Scores are computed from multiple evidence types; peak list may be incomplete in API response
Region search returns 0 results	Region is too small or on an uncharacterized chromosome	Use at least 10 kb windows; avoid alternative contigs (chrUn_, random_)

Related Skills

• gwas-database — NHGRI-EBI GWAS Catalog for published SNP-trait associations; use alongside RegulomeDB to prioritize GWAS hits
• clinvar-database — Clinical pathogenicity classifications; complements RegulomeDB's functional regulatory evidence
• encode-database — Direct ENCODE REST API access for TF ChIP-seq and ATAC-seq peak sets that underlie RegulomeDB scores
• ensembl-database — Variant annotation and gene coordinate lookup; use to map rsIDs to genomic positions

References

• RegulomeDB Official Help — Full documentation and scoring schema
• RegulomeDB API endpoint — Main REST API endpoint reference
• Boyle AP et al. "Annotation of functional variation in personal genomes using RegulomeDB." Genome Research 22(9): 1790–1797 (2012). https://doi.org/10.1101/gr.137323.112
• Dong S et al. "Annotating genomic variants and their functional effects using RegulomeDB 2.0." Bioinformatics 35(24): 5341–5343 (2019). https://doi.org/10.1093/bioinformatics/btz560