Experiment / E0NLIWVBETargeted / Cap-STARR-seq

Targeted modified STARR-seq assay of cancer-risk SNP fragments in HEK293T

Systematic identification of regulatory variants associated with cancer risk

Genome-derived approximately 450–500-bp fragments centered on 10,673 cancer-risk-associated SNPs (996 GWAS tag SNPs plus LD proxies) were captured from pooled DNA of 10 Chinese Han individuals and inserted downstream of the luciferase ORF in a modified pGL4.23-SCP1-ccdB self-transcribing reporter. The library was electroporated into HEK293T cells, and regulatory activity was quantified from RNA-output versus plasmid-input fragment counts in two replicates, including allele-specific changes.

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:0063
Reference genome: hg19
Design focus: Variant-focused
Region of interest: Not reported / not applicable

Perturbation & assay details

Basal / Untreated

Modified pGL4.23-SCP1-ccdB plasmid with the tested fragments acting as self-transcribing elements in the luciferase 3′ UTR. Genomic DNA from ten Chinese Han individuals was pooled, sonicated, size-selected at approximately 450–500 bp, and targeted by a custom Nimblegen/Roche capture array spanning 250 bp upstream and downstream of each selected SNP. Two library/transfection replicates were sequenced on an Illumina HiSeq 2500 with 250-bp paired-end reads; RNA output was poly(A)-selected, DNase-treated cDNA and input was plasmid DNA. Reads were mapped to hg19, and activity was calculated from RNA-output versus DNA-input fragment counts; allele ratios from the same fragments were used for allelic effects.

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

fragment_id: Internal identifier for the approximately 500-bp SNP-containing library fragment.
snp_id: dbSNP rsID for the SNP in the tested fragment.
chromosome: Chromosome label for the SNP on the hg19 assembly.
position_hg19: 1-based SNP coordinate on hg19.
tag_snp_id: GWAS tag SNP linked to the tested SNP by the study's LD selection.
cancer_trait: Cancer or cancer-related GWAS trait associated with the tag SNP.
dna_input_rep1_count: Raw plasmid-library fragment count for the second input replicate (GSM2470988).
rna_output_rep1_count: Raw RNA/cDNA output fragment count for the second transfection replicate (GSM2470986).
dna_input_rep2_count: Raw plasmid-library fragment count for the third input replicate (GSM2470989).
rna_output_rep2_count: Raw RNA/cDNA output fragment count for the third transfection replicate (GSM2470987).
activity_log2fc: DESeq2 log2 fold change of RNA-output fragment counts over DNA-input counts.
activity_fold_change: Derived linear fold change equal to 2 raised to activity_log2fc.
activity_pvalue: DESeq2 Wald-test p-value for the fragment activity change.
activity_fdr: Benjamini–Hochberg-adjusted DESeq2 p-value for fragment activity.
normalized_ref_input_count: Authors' normalized reference-allele count in the combined input library.
normalized_ref_output_count: Authors' normalized reference-allele count in the combined output library.
normalized_alt_input_count: Authors' normalized alternative-allele count in the combined input library.
normalized_alt_output_count: Authors' normalized alternative-allele count in the combined output library.
alt_vs_ref_odds_ratio: Odds ratio for the alternative/reference allele ratio in output versus input.
allelic_log2_odds_ratio: Derived log2 of alt_vs_ref_odds_ratio; NA when the source odds ratio is zero or unavailable.
allelic_pvalue: Two-sided Fisher exact-test p-value for the change in allele ratios.
allelic_fdr: Benjamini–Hochberg-adjusted p-value for the allele-specific test.
activity_class: Source activity label: PRE (positive regulatory element), NRE (negative regulatory element), or Inactive.
allele_test_qc_pass: TRUE when the source provides numeric allelic Fisher p-value and FDR; FALSE when the allele-specific count/QC threshold was not met or the test was unavailable.
reported_regulatory_variant: TRUE for PRE/NRE rows with source allelic FDR below 0.1, reproducing the paper's 70 reported regulatory SNPs.

Quality control

The authors mapped reads to hg19 with BWA 0.7.12-r1039, retained 400–600-bp fragments overlapping a selected SNP, and normalized counts with the DESeq2 median-of-ratios method. Of 10,673 designed SNP-containing regions, the processed table retains the 7,725 rows labeled PRE, NRE, or Inactive after the DESeq2 low-count filter; the 2,948 rows labeled NA were excluded. For allele-specific testing, pooled normalized reference/alternative counts with low coverage were excluded by the authors (count threshold 10), and two-sided Fisher exact-test p-values were Benjamini–Hochberg adjusted to FDR <0.1. Rows without a valid allelic p-value/FDR are flagged allele_test_qc_pass=FALSE rather than removed because their fragment-level activity passed QC.

Curation notes

The paper calls this a modified STARR-seq approach; it is classified here as Targeted / Cap-STARR-seq because genomic fragments were targeted by custom capture before cloning. Dataset S1 supplies fragment activity and allele metrics, and the four GEO processed fragment-count files were cross-checked against the spreadsheet counts for all 10,673 SNP IDs. Dataset S2 is dual-luciferase validation and is retained only in raw_data, not represented as a separate MPRA experiment. The authors note that the 3′ UTR reporter configuration can also capture post-transcriptional effects such as mRNA stability; only the single HEK293T basal condition was screened.

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