HepG2 STARR-seq allelic enhancer activity screen
Systematic analysis of binding of transcription factors to noncoding variantsA pooled human STARR-seq library tested 190-bp genomic fragments containing alleles of 3,943 noncoding SNPs in HepG2 cells, together with 37 known enhancer controls and 2,998 yeast ORF negative controls. Three biological replicates were generated, with two technical sequencing libraries per replicate.
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.
Perturbation & assay details
Basal / Untreated
Episomal human STARR-seq plasmids were transfected into HepG2 cells with Fugene HD and harvested 48 hours later. The library used SNP-centered and ATAC-summit-centered 190-bp genomic inserts with constant amplification/cloning flanks, plus known enhancer and yeast ORF controls; poly(A)+ reporter RNA was reverse-transcribed, amplified, and sequenced. GEO labels the two technical library types as ssIII and AB; these were summed within each of three biological replicates before processing.
Processed data
50 rows per page. Click a cell to inspect its full value.
Visible columns (42 of 42)
| 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 |
Filters apply to this table only. The CSV download contains the complete processed table; filtered rows are available through the API.
Column dictionary · 42 definitions
- oligo_id
- Exact oligo identifier from the GEO count matrix and Supplementary Table 5.
- element_id
- Context-specific element identifier; variant contexts distinguish SNP-centered and summit-centered inserts.
- variant_id
- Chromosome and 1-based SNP position in chr:position form; blank for controls.
- rs_id
- dbSNP rs identifier recovered by matching the coordinate to Supplementary Table 1.
- element_type
- SNP-containing genomic fragment, known enhancer positive control, or yeast ORF negative control.
- design_set
- Supplementary Table 5 library set: snpcenter, summitcenter, positive, or yeast.
- chromosome
- Chromosome or yeast ORF identifier from Supplementary Table 5.
- position
- Source position; human variant positions are 1-based, while yeast control positions retain the source midpoint.
- fragment_start
- Insert start coordinate parsed from the oligo identifier where supplied; blank for SNP-centered inserts without explicit bounds.
- fragment_end
- Insert end coordinate parsed from the oligo identifier where supplied; blank for SNP-centered inserts without explicit bounds.
- allele
- Allele represented by the oligo; NONE for non-allelic controls.
- reference_allele
- Reference allele from the author pbSNP or paSNP annotation when available.
- alternative_alleles
- Semicolon-separated alternative allele(s) from the author pbSNP or paSNP annotation.
- allele_role
- Reference, alternative, other, or unoriented relative to the available author allele annotation.
- reported_pbSNP
- TRUE when the SNP is listed in Supplementary Table 3 as a preferential-binding SNP; blank for controls.
- pbSNP_affected_TFs
- Semicolon-separated transcription factors listed as affected for the pbSNP in Supplementary Table 3.
- reported_paSNP
- TRUE when the coordinate is listed in Supplementary Table 6 as preferentially active in HepG2; blank for controls.
- paSNP_reference_allele
- Reference allele in the cell-specific Supplementary Table 6 paSNP call.
- paSNP_alternative_alleles
- Semicolon-separated alternative allele(s) in the cell-specific Supplementary Table 6 paSNP call.
- input_count
- Raw count in the synthesized plasmid input library.
- input_cpm
- Input count normalized to total input-library counts as counts per million.
- rna_count_rep1
- RNA reporter count for biological replicate 1 after summing ssIII and AB technical libraries.
- rna_count_rep2
- RNA reporter count for biological replicate 2 after summing ssIII and AB technical libraries.
- rna_count_rep3
- RNA reporter count for biological replicate 3 after summing ssIII and AB technical libraries.
- rna_cpm_rep1
- Replicate 1 RNA count normalized to the merged HepG2 RNA library total as counts per million.
- rna_cpm_rep2
- Replicate 2 RNA count normalized to the merged HepG2 RNA library total as counts per million.
- rna_cpm_rep3
- Replicate 3 RNA count normalized to the merged HepG2 RNA library total as counts per million.
- log2_activity_rep1
- Library-size-normalized log2 RNA/input activity for replicate 1, using a 0.5-count pseudocount.
- log2_activity_rep2
- Library-size-normalized log2 RNA/input activity for replicate 2, using a 0.5-count pseudocount.
- log2_activity_rep3
- Library-size-normalized log2 RNA/input activity for replicate 3, using a 0.5-count pseudocount.
- mean_log2_activity
- Mean of the three replicate log2 RNA/input activity scores.
- sd_log2_activity
- Sample standard deviation of the three replicate log2 activity scores.
- mean_activity_fold
- 2 raised to mean_log2_activity; normalized RNA/input activity on a fold scale.
- n_alleles_designed
- Number of designed oligo alleles in the same context-specific element before QC.
- n_alleles_retained
- Number of allele oligos retained for the same context-specific element after QC.
- allele_activity_range_log2
- Difference between the highest and lowest retained allele mean log2 activity in the element.
- most_active_allele
- Allele with the highest retained mean log2 activity in the element.
- least_active_allele
- Allele with the lowest retained mean log2 activity in the element.
- allele_effect_vs_reference_log2
- Agent-derived mean alternative-minus-reference log2 activity across three replicates for this alternative allele.
- agent_paired_t_pvalue_vs_reference
- Agent-derived two-sided paired t-test p-value across three replicate activity differences versus the reference allele; not the paper's limma p-value.
- agent_paired_t_fdr_vs_reference
- Benjamini-Hochberg FDR for the agent-derived paired tests across retained HepG2 element/allele comparisons.
- qc_pass
- TRUE for every row retained after the stated input and merged-replicate coverage filter; failed rows are omitted.
Quality control
The paper's STARR-seq QC required input coverage >25 reads and >5 reads in at least three libraries after technical replicate handling, and used yeast oligos to estimate common dispersion for edgeR enrichment testing. Here ssIII and AB counts were summed within each biological replicate, then the stated filter was applied: 14,932 of 14,996 oligos passed and 64 were removed (57 failed the input threshold; 37 failed the coverage threshold; 64 failed at least one threshold). The paper's reported paSNP calls (Supplementary Table 6; author FDR <0.01 from paired limma testing) are retained as annotations. The processed table also contains an agent-derived exploratory paired t-test/FDR across three replicate log2 activity scores; these values are not the authors' edgeR/limma statistics and were not used for filtering.
Curation notes
This child experiment packages the paper's genuine STARR-seq validation rather than the study's primary SNP-SELEX binding measurements. Hep-G2 was resolved to Cellosaurus CVCL:0027. Supplementary Table 6 contains 200 HepG2 rows representing 198 distinct coordinates after collapsing duplicate allele entries; all 198 coordinates remain represented in the QC-passing table. The paSNP flag is repeated for every retained allele/context row at a reported coordinate, including both SNP-centered and summit-centered constructs when present. The processed table omits the long insert sequence to keep the analysis table compact; the exact sequences remain in raw_data/Supplementary_Table_5_STARR_seq_oligos.xlsx and its CSV conversion. The calculated activity scores are library-size-normalized summaries of the supplied counts and should not be treated as a reimplementation of the authors' edgeR model output.