Experiment / E7WYY875ZPromoter / Core Promoter MPRA

Reg-Seq promoter MPRA across E. coli growth conditions and transcription-factor knockout validations

Deciphering the regulatory genome of Escherichia coli, one hundred promoters at a time

A library of randomly mutagenized 160-bp E. coli promoter regions was cloned upstream of GFP on the episomal pSC101-derived pJK14 plasmid and linked to random 20-nt genetic barcodes. RNA-seq barcode expression measurements were compared with DNA-library counts across growth conditions and knockout backgrounds; table.csv packages the released signed information footprints at single-base resolution.

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

Multiple carbon-source, stress, and oxygen conditions plus transcription-factor knockout validations; the exact released condition code and expanded description are in table.csv.

Episomal pSC101/pJK14 reporter assay in E. coli K-12 MG1655-derived backgrounds. Each promoter region was 160 bp (115 bp upstream and 45 bp downstream of the presumed TSS), with an average mutation rate of about 10%, approximately 2,200 unique promoter sequences per gene, and an average of five unique 20-nt barcodes per variant promoter. The construct contains the barcode, RBS, GFP coding sequence, and terminator. The released information footprints are signed mutual-information-derived outputs from DNA-versus-mRNA barcode counts; the paper also used inferred energy matrices and mass spectrometry for regulatory-factor support.

Processed data

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

Visible columns (15 of 15)
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 · 15 definitions
gene
Gene or operon label for the released promoter construct.
condition_code
Condition or knockout code parsed from the released information-footprint filename.
condition_class
Normalized class: growth_condition, tf_knockout, or auxiliary_or_unresolved.
condition_description
Human-readable condition description; unresolved released codes are explicitly identified.
tss_coordinate
Genomic coordinate of the presumed transcription start site from the released wild-type promoter design table.
transcription_direction
Forward or reverse transcription orientation for the promoter design.
position_0based
Zero-based position within the 160-bp mutagenized promoter region.
position_relative_tss
Promoter position relative to the presumed TSS, from -115 through +44 bp.
wild_type_base
Wild-type nucleotide at the corresponding promoter position.
info_signed
Released signed information-footprint value in bits; the sign records the direction of expression change associated with mutation under the Reg-Seq convention.
info_magnitude_bits
Absolute value of the released information-footprint value in bits.
mutation_effect_direction
Interpretation of the signed footprint: repressor_like means mutation increases expression, activator_like means mutation decreases expression, and no_direction is exactly zero.
reported_binding_site_type
Curated binding-site type overlapping this position from eLife Supplementary file 3, such as activator or repressor; blank means no reported overlap.
reported_binding_site_identity
Curated transcription-factor identity overlapping this position when supplied by Supplementary file 3; blank means none was supplied.
source_file
Relative path to the released raw information-footprint file used for the row.

Quality control

Author QC included retaining promoter libraries with an average mutation rate of 9.5–10.5%; removing mapping barcodes linked to insertions/deletions, multiple promoter variants, or singleton observations; filtering reads with PHRED <20; discarding cDNA samples with DNA contamination at or above 5% of the mRNA concentration; and not analyzing gene-condition datasets with fewer than 20,000 usable reads. Experiments used two biological replicates. Package-level QC retained footprint files only when the released file contained finite numeric values at every position 0–159 and the gene resolved to a released 160-bp wild-type sequence; 1,239 files passed and ten unmatched comparison files were excluded. Non-significant positions remain in the table because they are valid measurements, not QC failures.

Curation notes

This is one promoter-library MPRA experiment represented across multiple conditions, not a separate experiment per condition. The table is a position-level derived footprint resource rather than a raw FASTQ or full barcode-count matrix; a compact ykgE barcode/count example is retained under raw_data. The paper describes 113 TSS/promoters and its abstract describes 12 growth conditions, while the released wild-type design table and footprint filenames contain 107 matched promoter records, 13 normalized growth-condition codes including acetate and xanth2, eight knockout codes, and one auxiliary fum1 code. The condition codes are preserved verbatim and only the explicit mappings supported by the paper or the authors’ interactive resource are expanded. The no-cAMP condition used the JK10 background; the biosample CURIE therefore describes the principal MG1655 background rather than every validation strain. Ten released footprint files for comparison genes could not be joined to a 160-bp wild-type design record and were excluded from table.csv rather than assigned speculative sequence 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.