Experiment / E2F7DFQM1Standard STARR-seq

Population-scale STARR-seq assay of allele-specific regulatory activity

Massively parallel quantification of the regulatory effects of noncoding genetic variation in a human cohort

A targeted population STARR-seq library cloned 174 candidate regulatory amplicons from 95 HAPO-cohort donor genomes into the 3′ UTR of an episomal STARR-seq reporter. Seven HepG2 RNA output replicates were compared with the input plasmid DNA library to quantify allele-specific activity at 283 SNPs; phased haplotype effects are retained in raw_data.

Download processed table (CSV)Raw source data API documentation

Processed tables are specific to each experiment. Column names, units, measurements, and table structure are not standardized across the database. Check this experiment’s column definitions and quality-control notes before comparing or combining data.

Organism: Human
Taxonomy ID: NCBITaxon:9606
Biosample: CVCL:0027
Reference genome: hg19
Design focus: Variant-focused
Region of interest: chr3:156781820-157035990

Perturbation & assay details

Basal / Untreated

Population-scale targeted STARR-seq (POP-STARR-seq) adaptation. Multiplex-PCR-captured genomic amplicons from 95 donors were cloned into the 3′ UTR of an episomal reporter, expanded in E. coli, transiently transfected into HepG2 cells, and read by 250-bp paired-end sequencing of input plasmid DNA and expressed reporter mRNA. Seven independent RNA output replicates plus one input-library sample are represented in GEO; the paper's Methods also state eight replicate transfections, so replicate counts are noted as reported.

Processed data

50 rows per page. Click a cell to inspect its full value.

Visible columns (25 of 25)

Row
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50

Page 1 · 50 rows · More results available

Filters apply to this table only. The CSV download contains the complete processed table; filtered rows are available through the API.

Column dictionary · 25 definitions

variant_id: dbSNP rs identifier when deposited, otherwise a stable chromosome:position:allele0>allele1 identifier.
chromosome: Chromosome reported in Supplemental Data 4.
position_hg19_1based: 1-based genomic position on the hg19 reference assembly.
allele0_ref: Source VCF allele0/reference allele used for the allele-specific comparison.
allele1_alt: Source VCF allele1/alternate allele used for the allele-specific comparison.
dna_count_allele0: Pooled input-plasmid DNA read count assigned to allele0.
rna_count_allele0: Pooled expressed reporter mRNA read count assigned to allele0.
dna_count_allele1: Pooled input-plasmid DNA read count assigned to allele1.
rna_count_allele1: Pooled expressed reporter mRNA read count assigned to allele1.
dna_total: Derived total input DNA reads across allele0 and allele1.
rna_total: Derived total pooled reporter RNA reads across allele0 and allele1.
normalized_activity_allele0: Derived allele0 activity, (RNA allele0 / total RNA) / (DNA allele0 / total DNA).
normalized_activity_allele1: Derived allele1 activity, (RNA allele1 / total RNA) / (DNA allele1 / total DNA).
effect_ratio_allele1_vs_allele0: Deposited Supplemental Data 4 effect ratio, (RNA1/DNA1) / (RNA0/DNA0), so values above 1 favor allele1.
log2_effect_allele1_vs_allele0: Base-2 logarithm of the deposited allele1-versus-allele0 effect ratio.
higher_activity_allele: Allele with the higher derived normalized activity; equal denotes a tie.
raw_p_value: Deposited two-tailed Fisher exact-test P value for the allele-specific DNA/RNA comparison.
fdr_adjusted_p_value: Benjamini-Hochberg adjusted P value recomputed across all 283 deposited SNP P values.
significant_fdr_0_05: True when the recomputed Benjamini-Hochberg adjusted P value is <0.05.
source_p_lt_0_05: True when the deposited source P value is <0.05; this preserves the source threshold that reproduces the paper's reported 36-variant set and is not relabeled as recomputed FDR.
allele_frequency_class: Rare when capture MAF is <1%; otherwise common.
maf_capture: Minor allele frequency in the capture cohort, from Supplemental Data 4.
maf_pop_starr: Minor allele frequency in the population STARR-seq reporter library, from Supplemental Data 4.
amplicon_ids: Semicolon-separated Supplemental Data 8 amplicon IDs whose inclusive haplotype-effect interval contains the variant position; blank means no interval match.
amplicon_has_significant_haplotype: True when any matched amplicon has a Supplemental Data 8 haplotype Padj <0.05; false when matched but none is significant; blank means no interval match.

Quality control

The authors hard-filtered GATK variant calls (SNP thresholds: QD >=2, MQ >=40.0, FS <=60.0, MQRankSum >=-12.5, ReadPosRankSum >=-8.0, QUAL >=100.0; indel thresholds: QD >=2, FS <=200.0, ReadPosRankSum >=-20.0, InbreedingCoeff >=-0.8, QUAL >=100.0), phased with SHAPEIT2 read-aware phasing, aligned reads to imputed haplotypes with strict Bowtie2 mismatch/gap penalties of 100, pooled replicate reads, discarded variants with fewer than two total reads in input DNA or pooled RNA, and used Fisher's exact test with pseudocount 1 followed by Benjamini-Hochberg FDR. The package retains the 283 source-tested SNPs because every row has total input DNA and pooled RNA counts >2; no rows failed this depth QC. The table recomputes BH FDR (23 q<0.05) and also records the source P<0.05 set (36 rows, matching the paper's stated 27 common + 9 rare), because the deposited P column and the paper's FDR count are internally inconsistent.

Curation notes

The study's primary MPRA experiment is the population-scale STARR-seq library; luciferase validations are not packaged as a separate experiment. The table is variant-centered and contains all 283 source-tested SNPs. The paper reports 36 variants at FDR<0.05 (27 common, 9 rare), but the deposited P values yield 36 at nominal P<0.05 and 23 after an independent BH adjustment across all 283 tests; both flags are preserved. Allele0/allele1 are source VCF alleles, and the source effect is allele1/allele0, so rs4266144's higher-activity C allele is represented by an effect ratio below 1 because C is allele0. Supplemental Data 4, 7, and 8 were reconstructed as TSV tables from the NCBI BioC supplementary representation; the original BioC response is retained in raw_data for provenance. The source includes five individual allele count cells below 2, but their total DNA/RNA library depths exceed 2 and the original 283-row source table retains them; this package follows the stated total-library depth interpretation. The article's Results describe seven independent transfections while Methods says eight; GEO contains seven RNA output samples plus one input plasmid sample.

Row
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50

Row
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50

Row
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50