Experiment / E6ZN1BFLJIntegrated lentiMPRA

Integrated lentiMPRA of duplicated CRE tiles in GM12878

Influence of cis-regulatory elements on expression divergence in human segmental duplications

A shared library of 200-bp tiles from 26 human-specific segmental-duplication gene families, including human ancestral/derived paralog and panTro6 chimpanzee homolog sequences plus scrambled negative controls, was delivered by lentivirus to the GM12878 lymphoblastoid cell line. Three technical replicates were quantified from integrated DNA and expressed RNA barcodes with MPRAnalyze; the processed table retains 7,750 construct-level test measurements that pass the available construct and result QC.

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:7526
Reference genome: GRCh38
Design focus: Region-focused
Region of interest: Not reported / not applicable

Perturbation & assay details

Basal / Untreated

Candidate CREs were identified from remapped H3K27ac, ATAC-seq, and DNase-seq peaks in ancestral HSD loci and tiled at 200 bp with 100-bp overlap. Homologous human GRCh38 and chimpanzee panTro6 sequences were synthesized with universal pLS-SceI priming sequences, cloned into a lentiviral reporter, and barcode-associated with MPRAflow. GM12878 cells were transduced at approximately MOI 13 in RPMI with 10% FBS and 8 micrograms/mL polybrene; nucleic acids were extracted 48 hours after infection. Barcode association used a PE150 NextSeq run and DNA/RNA barcode counting used three PE15 NextSeq runs. Activity is the MPRAnalyze estimated transcription rate alpha from integrated DNA and expressed RNA counts.

Processed data

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

Visible columns (28 of 28)

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 · 28 definitions

element_id: Unique published construct/sequence name.
ancestral_coord_grch38: GRCh38 coordinate of the ancestral HSD tile used as the sequence-family anchor.
homolog_coord: One or more homologous coordinates represented by the construct, from Supplemental Table S6; coordinates may be human GRCh38 or chimpanzee panTro6 and are pipe-delimited when shared.
homolog_species: Species represented by the homologous coordinate(s): human, chimpanzee, or human;chimpanzee for a construct shared across species.
homolog_class: Coordinate class relative to the ancestral anchor: human_ancestral, human_derived, chimpanzee_ortholog, or combinations joined with +.
ancestral_coord_strand: Strand of the ancestral GRCh38 coordinate as supplied in Supplemental Table S6.
nearest_feature_strand: Strand assignment from the nearest annotated feature as supplied in Supplemental Table S6.
alignment_size_difference_bp: Difference in original homologous alignment size from the 200-bp target, before trimming or padding.
sequence_200bp: The synthesized 200-bp regulatory sequence for the construct, without the universal priming flanks.
grch38_t2t_chm13_comparison: Supplemental Table S6 comparison of the sequence between GRCh38 and T2T-CHM13, such as identical, different, mixed, or NA.
pantro6_mpantro3_comparison: Supplemental Table S6 comparison of the chimpanzee sequence between panTro6 and mPanTro3 haplotype 1.
construct_type: Published MPRA quantification type; retained rows are test constructs.
target_coordinates: Target coordinate field from the authors' MPRA quantification output.
ancestor_locus_from_quant: Ancestral tile/locus field from the authors' MPRA quantification output.
alpha_activity: MPRAnalyze estimated transcription rate alpha, the primary construct activity measure modeled from DNA and RNA barcode counts.
zscore: MPRAnalyze z-score activity statistic from the published quantification output.
pval_zscore: P-value associated with the z-score activity statistic.
mad_score: Median-absolute-deviation activity score used for testing against scrambled negative controls.
pval_mad: P-value for the median-absolute-deviation activity test against scrambled negative controls.
activity_qvalue: Author-supplied qval field from the MPRA quantification export, retained as provided.
active_vs_scramble: Derived Boolean activity call: true when pval_mad < 0.05, matching the paper's active-sequence threshold.
log2fc_vs_chimpanzee: Published pairwise log2 fold change for this construct relative to the chimpanzee baseline; blank when no exact Supplemental Table S10 row is available.
pval_vs_chimpanzee: Published pairwise p-value for the construct-versus-chimpanzee homolog test.
fdr_vs_chimpanzee: Published FDR for the construct-versus-chimpanzee homolog test.
log2fc_vs_human_ancestral: Published pairwise log2 fold change for this construct relative to the human ancestral homolog; blank when no exact Supplemental Table S12 row is available.
pval_vs_human_ancestral: Published pairwise p-value for the construct-versus-human-ancestral homolog test.
fdr_vs_human_ancestral: Published FDR for the construct-versus-human-ancestral homolog test.
qc_pass: Boolean indicating that the row passed the package-level construct/result QC filter and was retained.

Quality control

The authors used three technical replicates, MPRAflow barcode-insert association with map quality 1 and minimum barcode coverage 10, manual review of promiscuous barcode assignments, total-sum depth normalization, and MPRAnalyze models. Within-cell-type replicate reproducibility was reported as mean pairwise r = 0.93 for both DNA and RNA libraries. The 500 scrambled ancestral tiles served as negative controls; activity was called by the authors at pval.mad < 0.05, yielding 1,850 active sequences in GM12878. For this package, only published quantification rows with type = test, complete numeric activity fields, an exact 200-bp sequence, and a matching Supplemental Table S6 annotation were retained. The source quantification contains 7,750 test constructs after the authors' barcode/count filtering; inactive but technically valid constructs remain in the table and are marked by active_vs_scramble = false.

Curation notes

This child experiment represents the GM12878 condition of the shared integrated lentiMPRA library. The table was assembled from the published Supplemental Table S6 sequence annotations, Supplemental Table S8 GM12878 activity results, the authors' lcl_quant.txt export, and Supplemental Table S10/S12 pairwise result sheets. It is construct-level rather than barcode-level and does not include raw FASTQ. Multi-target names and coordinates are retained verbatim; combined homolog_class values indicate one construct mapped to multiple homologs. The paper notes that both host cell lines are cancer-derived or have abnormal karyotypes. The parent raw_data folder preserves the full supplement, MPRA code/data archive, and ENA run metadata.

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