Experiment / E65D3R0I0Whole-Genome STARR-seq (WHG-STARR-seq)

Whole-genome STARR-seq enhancer activity in 2iL- and serum/LIF mouse embryonic stem cells

STARR-seq identifies active, chromatin-masked, and dormant enhancers in pluripotent mouse embryonic stem cells

An episomal whole-genome STARR-seq library tested randomly sheared mouse genomic fragments in E14Tg2a embryonic stem cells cultured in ground-state 2i+LIF (2iL) or metastable serum+LIF (SL). The experiment used two biological replicates per condition with matched input-DNA libraries, and the table preserves the union of source-defined STARR loci plus the active-chromatin/no-STARR C3 comparison class.

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

2i+LIF (ground-state) versus serum+LIF (metastable) culture conditions; no additional treatment

Mouse genomic DNA was sonicated and size-selected at approximately 700–1200 bp (median approximately 850 bp), adapter-ligated, and cloned downstream of the minimal SCP1 promoter in the hSTARR-seq_SCP1 plasmid (Addgene 99292). Four hundred micrograms of plasmid pool were transfected into approximately 400 million E14Tg2a cells with Lipo3000; cells were harvested after 24 hours. PolyA+ reporter RNA was sequenced as four STARR output libraries and transfected plasmid DNA as four matched input libraries: 2iL biological replicates 1–2 and SL biological replicates 1–2. Libraries were paired-end sequenced, mapped to mm9 with BWA-MEM, and quantified as RNA/input enrichment.

Processed data

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

Visible columns (48 of 48)
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 · 48 definitions
element_id
Unique STARR-seq locus identifier from the author table.
chromosome
Reported mouse chromosome for the locus.
start
Reported locus start coordinate from the source mm9 table; coordinate convention is retained as supplied.
end
Reported locus end coordinate from the source mm9 table; coordinate convention is retained as supplied.
length_bp
Source-reported locus length in base pairs.
summit
Source-reported STARR-seq peak summit coordinate.
feature
Source genomic feature annotation, such as Intergenic or Intron.
chromatin_class
Author-defined locus class: C1, STARR-seq plus active chromatin; C2, STARR-seq without the active chromatin signature; C3, active chromatin without significant STARR-seq signal.
starr_active_2iL
Boolean derived from source enrichment_2iL >= 3 and padj_2iL1 and padj_2iL2 < 0.05.
starr_active_SL
Boolean derived from source enrichment_SL >= 3 and padj_SL1 and padj_SL2 < 0.05.
starr_active_any
Boolean indicating source-defined STARR activity in either 2iL or SL.
rna_2iL_rep1
Total STARR-seq reporter-RNA read count for 2iL biological replicate 1.
rna_2iL_rep2
Total STARR-seq reporter-RNA read count for 2iL biological replicate 2.
rna_SL_rep1
Total STARR-seq reporter-RNA read count for SL biological replicate 1.
rna_SL_rep2
Total STARR-seq reporter-RNA read count for SL biological replicate 2.
unique_rna_2iL_rep1
Non-duplicated STARR-seq reporter-RNA read count for 2iL replicate 1.
unique_rna_2iL_rep2
Non-duplicated STARR-seq reporter-RNA read count for 2iL replicate 2.
unique_rna_SL_rep1
Non-duplicated STARR-seq reporter-RNA read count for SL replicate 1.
unique_rna_SL_rep2
Non-duplicated STARR-seq reporter-RNA read count for SL replicate 2.
input_dna_2iL_rep1
Matched transfected-plasmid input-DNA read count for 2iL replicate 1.
input_dna_2iL_rep2
Matched transfected-plasmid input-DNA read count for 2iL replicate 2.
input_dna_SL_rep1
Matched transfected-plasmid input-DNA read count for SL replicate 1.
input_dna_SL_rep2
Matched transfected-plasmid input-DNA read count for SL replicate 2.
enrichment_2iL_rep1
Source STARR-seq-over-input enrichment for 2iL replicate 1, with source shrinkage/normalization applied.
enrichment_2iL_rep2
Source STARR-seq-over-input enrichment for 2iL replicate 2, with source shrinkage/normalization applied.
enrichment_2iL_merged
Source condition-level 2iL enrichment after merging the two biological replicates.
enrichment_SL_rep1
Source STARR-seq-over-input enrichment for SL replicate 1, with source shrinkage/normalization applied.
enrichment_SL_rep2
Source STARR-seq-over-input enrichment for SL replicate 2, with source shrinkage/normalization applied.
enrichment_SL_merged
Source condition-level SL enrichment after merging the two biological replicates.
padj_2iL_rep1
Benjamini–Hochberg-adjusted binomial p-value for the 2iL replicate 1 locus test.
padj_2iL_rep2
Benjamini–Hochberg-adjusted binomial p-value for the 2iL replicate 2 locus test.
padj_SL_rep1
Benjamini–Hochberg-adjusted binomial p-value for the SL replicate 1 locus test.
padj_SL_rep2
Benjamini–Hochberg-adjusted binomial p-value for the SL replicate 2 locus test.
log2fc_SL_vs_2iL
Author-reported DESeq2 log2 fold change for SL versus 2iL STARR-seq activity; blank for C3 loci without a differential STARR signal.
pval_SL_vs_2iL
GEO-supplied DESeq2 p-value for differential STARR-seq activity between SL and 2iL; blank for C3 loci.
padj_SL_vs_2iL
GEO-supplied Benjamini–Hochberg-adjusted DESeq2 p-value for SL versus 2iL; blank for C3 loci.
atac_signal_2iL
Author-provided ATAC-seq signal associated with the locus in 2iL cells.
atac_signal_SL
Author-provided ATAC-seq signal associated with the locus in SL cells.
p300_signal_union
Author-provided P300 overlap/call in the union chromatin annotation.
med1_signal_2iL
Author-provided MED1 ChIP-seq signal associated with the locus in 2iL cells.
med1_signal_SL
Author-provided MED1 ChIP-seq signal associated with the locus in SL cells.
p300_signal_2iL
Author-provided P300 ChIP-seq signal associated with the locus in 2iL cells.
p300_signal_SL
Author-provided P300 ChIP-seq signal associated with the locus in SL cells.
gc_fraction_center_500bp
GC fraction calculated by the authors in the central 500 bp of the locus.
cpg_count_center_500bp
Number of CpGs calculated by the authors in the central 500 bp of the locus.
epicseg_state_2iL
Numeric EpiCSeg five-state chromatin segmentation code for 2iL; state labels/intervals are supplied in raw_data/author_Epistates.tsv.
epicseg_state_SL
Numeric EpiCSeg five-state chromatin segmentation code for SL; state labels/intervals are supplied in raw_data/author_Epistates.tsv.
package_qc_pass
TRUE for rows retained after package-level ID, interval, cross-source, and finite-value integrity checks.

Quality control

The paper discarded mapped reads with MAPQ < 10 and non-proper pairing, and assessed library complexity, GC bias, insert size, PCR duplication, and genome coverage with PICARD and bedtools. Initial loci were called with MACS2 against matched input DNA, read counts were obtained with featureCounts, per-library significance used a right-tailed binomial model with Benjamini–Hochberg correction, and enrichment was stabilized with beta-binomial Bayesian shrinkage. One million GC- and size-matched random regions supported the empirical enrichment threshold. Final STARR loci were defined by condition-level enrichment >= 3 and adjusted binomial p < 0.05 in both biological replicates. Differential activity used DESeq2 custom normalization for RNA/input depth and local input coverage; absolute fold change >= 2.5 and p < 0.05 defined differential loci. Package QC required unique locus IDs, agreement between the author and GEO tables for chromosome/interval/class, finite numeric values, nonnegative starts, end > start, and summit within the reported interval. One source row (starr_15915) was excluded because its reported summit (52967375) falls outside chr6:52967360-52967369. The processed table retains 27565 of 27566 source rows; C3 rows are intentionally retained as the authors' active-chromatin/no-STARR dormant comparison class rather than treated as failed sequencing rows.

Curation notes

The study's MPRA is a whole-genome episomal STARR-seq assay, represented by one experiment folder because the eight sequencing libraries are four RNA/input pairs spanning two culture conditions analyzed as one comparison. The source union has 18544 C1, 7072 C2, and 1950 C3 loci; the processed table retains 18543 C1, 7072 C2, and 1950 C3 rows after removing the single anomalous-summit row. Source-defined active counts are 18116 loci in 2iL and 18543 in SL; the packaged table has 18116 and 18542 respectively because the excluded row was SL-active. C3 is retained to preserve the paper's dormant/negative chromatin comparison, and its differential fields are null because it was not part of the STARR-positive set. E14Tg2a is resolved to Cellosaurus CVCL:9108 (ES-E14TG2a; 129P2/Ola mouse embryonic stem cell). Raw FASTQ/BW archives are not packaged; the raw-read GEO archive remains linked in the parent metadata.

Cite OpenMPRA

Cite the OpenMPRA database. Include your access date because the collection changes over time.

Please also cite the source studies when using their data.