Experiment / E008VW64ZAAV-MPRA / in vivo MPRA

HaCaT PGK AAV-STARR-CRAAVT enhancer screen

STARR-CRAAVT: A platform to identify cell type-specific regulatory elements for next-generation gene therapy

The PGK AAV-STARR-CRAAVT candidate library was transduced into HaCaT cells with AAV6.2 across three replicate wells at MOI 100,000 viral genomes per cell. Matched cellular DNA and RNA/cDNA counts quantify candidate representation and self-transcribed regulatory activity after 72 h.

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:0038
Reference genome: GRCh38
Design focus: Region-focused
Region of interest: Not reported / not applicable

Perturbation & assay details

Basal / Untreated

STARR-CRAAVT captured candidate-containing fragments from sheared human genomic DNA with Agilent SureSelectXT, cloned them into an AAV-STARR-seq transfer plasmid, and packaged the PGK library in AAV6.2. HaCaT cells were transduced at MOI 100,000 viral genomes per cell and harvested after 72 h; matched cellular DNA and poly(A)-enriched RNA/cDNA were sequenced as paired-end 76+8+8+76 reads on an Illumina NextSeq 500. Reads were mapped to hg38, uniquely mapped target-region fragments were counted, and RNA/DNA activity was used as the enhancer-strength readout.

Processed data

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

Visible columns (32 of 32)

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 · 32 definitions

element_id: Candidate genomic fragment identifier from the raw GEO Genomic_Region field; coordinates and sequence were not deposited in this matrix.
normalized_matrix_element_id: Identifier used for the matching row in the GEO normalized matrix; two minP/4.26 tile IDs differ between the raw and normalized files and are explicitly mapped here.
cell_type: Human cell line used for the cellular DNA/RNA screen.
promoter: STARR-CRAAVT promoter in the library: 32-bp minimal TATA-box promoter (minP), 276-bp alpha-fetoprotein promoter (AFP), or 511-bp phosphoglycerate kinase promoter (PGK).
delivery_mode: Construct delivery condition.
input_library: Input count source used for the library-level raw input column.
input_raw_count: Raw featureCounts read count for the corresponding plasmid or AAV input library.
dna_raw_rep1: Raw featureCounts count from cellular DNA replicate 1; used to assess candidate library representation.
rna_raw_rep1: Raw featureCounts count from cellular RNA/cDNA replicate 1; self-transcribed reporter output.
dna_normalized_rep1: GEO-supplied normalized cellular DNA count for replicate 1.
rna_normalized_rep1: GEO-supplied normalized cellular RNA/cDNA count for replicate 1.
dna_raw_rep2: Raw featureCounts count from cellular DNA replicate 2; used to assess candidate library representation.
rna_raw_rep2: Raw featureCounts count from cellular RNA/cDNA replicate 2; self-transcribed reporter output.
dna_normalized_rep2: GEO-supplied normalized cellular DNA count for replicate 2.
rna_normalized_rep2: GEO-supplied normalized cellular RNA/cDNA count for replicate 2.
dna_raw_rep3: Raw featureCounts count from cellular DNA replicate 3; used to assess candidate library representation.
rna_raw_rep3: Raw featureCounts count from cellular RNA/cDNA replicate 3; self-transcribed reporter output.
dna_normalized_rep3: GEO-supplied normalized cellular DNA count for replicate 3.
rna_normalized_rep3: GEO-supplied normalized cellular RNA/cDNA count for replicate 3.
dna_detected_reps: Number of eligible DNA replicates with raw count > 0; package QC required at least 2.
rna_detected_reps: Number of eligible RNA/cDNA replicates with raw count > 0; reported descriptively and not used as a filter because zero RNA can represent low or repressive activity.
dna_mean_raw_count: Arithmetic mean raw cellular DNA count across eligible replicates.
rna_mean_raw_count: Arithmetic mean raw cellular RNA/cDNA count across eligible replicates.
dna_mean_normalized_count: Arithmetic mean of GEO-supplied normalized cellular DNA counts.
rna_mean_normalized_count: Arithmetic mean of GEO-supplied normalized cellular RNA/cDNA counts.
log2_activity_rep1: Package-derived replicate activity score log2((normalized RNA/cDNA + 1) / (normalized DNA + 1)) for replicate 1; the +1 pseudocount avoids undefined values at zero.
log2_activity_rep2: Package-derived replicate activity score log2((normalized RNA/cDNA + 1) / (normalized DNA + 1)) for replicate 2; the +1 pseudocount avoids undefined values at zero.
log2_activity_rep3: Package-derived replicate activity score log2((normalized RNA/cDNA + 1) / (normalized DNA + 1)) for replicate 3; the +1 pseudocount avoids undefined values at zero.
log2_activity_mean: Arithmetic mean of the replicate-level package-derived log2 activity scores; a proxy for the paper-described RNA/DNA log2FC because GEO did not provide per-element DESeq2 results.
log2_activity_sd: Sample standard deviation of the replicate-level package-derived log2 activity scores.
activity_class: Descriptive class: enhancer_like if mean activity > 1, repressive_like if < -1, otherwise neutral_or_weak.
study_enhancer_threshold: Boolean flag for mean package-derived activity > 1, matching the paper’s nominal enhancer activity cutoff but not its adjusted-p-value requirement.

Quality control

The authors mapped paired-end reads to hg38 with bowtie2 v2.2.9, retained uniquely mapped reads with samtools, quantified reads overlapping capture-library target regions with featureCounts, and used MAPQ ≥30 for fragment-level analyses. Their comparative analysis removed candidates detected in only one replicate before pooling. Package QC retained 8821 of 15379 candidate rows by requiring raw cellular DNA count >0 in at least 2 of 3 available replicate wells and finite GEO-normalized cellular DNA/RNA values. RNA-positive filtering was not applied because zero reporter RNA is biologically interpretable as low or repressive activity. The paper’s enhancer definition also required RNA/DNA log2FC >1 and adjusted p<0.05; GEO released count matrices but not per-element DESeq2 p-values, so the table’s study_enhancer_threshold flag is only the nominal >1 activity cutoff and does not claim statistical significance.

Curation notes

Processed from GSE298485 raw and normalized count matrices. GEO sample accessions used: cellular DNA GSM9015205, GSM9015206, GSM9015207; cellular RNA/cDNA GSM9015228, GSM9015229, GSM9015230. The raw matrix contains 15,379 candidate fragment identifiers but no genomic coordinates or sequences, so element_id is retained exactly as deposited. The normalized matrix uses HCConly_minP_tile300/1100 where the raw matrix uses HCConly_426_tile300/1100; normalized_matrix_element_id preserves that alias mapping. Package-derived activity is the mean of replicate log2((normalized RNA/cDNA +1)/(normalized DNA +1)) scores; it is provided for consistent cross-condition comparison and should not be confused with the authors’ DESeq2 log2FoldChange. QC-retained rows are listed in source order.

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