Full-promoter Sort-Seq libraries were built from the three natural E. coli lac operators (O1, O2, and O3) with approximately 10% mutagenesis across the lacUV5 promoter and operator. Libraries were assayed in LacI copy-number backgrounds of approximately R=30, 62, 130, and 610 tetramers; the table summarizes published matrix-derived substitutions and appends the paper’s independent operator-mutant validation records.
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.
episomal pUA66 operator-GFP reporter; lacUV5 promoter and operator were mutagenized at approximately 10%; four FACS fluorescence bins were sequenced by Illumina; single-point Bayesian energy matrices were reported in kBT for primary R=130 matrices and in arbitrary units for several other contexts.
Processed data
50 rows per page. Click a cell to inspect its full value.
Visible columns (29 of 29)
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 · 29 definitions
record_type
Whether the row is an MPRA matrix-derived single-base substitution or a published flow-cytometry validation record.
model_id
Source S1 Models filename or compact repository summary filename.
library_design
Reporter-library design used for the matrix.
regulatory_context
Position of the LacI operator relative to the RNAP binding site.
reference_operator
Wild-type operator sequence used to design the library.
matrix_operator
Operator whose Sort-Seq matrix was used for the prediction.
reference_sequence
Reference DNA sequence for the matrix/library.
variant_id
Stable descriptive identifier for the substitution or validation mutant.
variant_sequence
21-bp operator sequence carrying the indicated substitution(s).
mutation_count
Number of sequence differences from the relevant reference sequence.
position_1based
1-based operator position for single-base substitutions; blank for multi-base validation records.
reference_base
Reference nucleotide at position_1based.
alternate_base
Alternative nucleotide at position_1based.
lacI_variant
LacI protein background used for the model.
lacI_copy_number_tetramers
Approximate mean LacI tetramer copy number R in the strain.
biological_day
Biological collection day when separately reported by the source model.
replicate_count
Number of split or biological replicate matrices summarized in the mean/SD.
matrix_energy_unit
Energy unit reported by the source matrix: kBT or arbitrary units (au).
delta_binding_energy_mean
Mean predicted binding-energy change for the alternative base relative to the matrix reference base; negative means stronger binding.
delta_binding_energy_sd
Standard deviation of the predicted relative binding-energy change across source matrix replicates.
Maximum replicate predicted relative binding-energy change.
predicted_binding_energy_kBT
Absolute predicted LacI binding energy in kBT when the source provides a kBT-scaled R=130 matrix.
predicted_binding_energy_sd_kBT
SD of the absolute kBT prediction across matrix replicates.
published_prediction_mean_kBT
Published mean binding-energy prediction from the repository validation summary.
published_prediction_sd_kBT
Published SD of the validation prediction.
published_measured_binding_energy_kBT
Published binding energy measured by fitting conventional flow-cytometry fold-change curves.
effect_direction
Interpretation of delta binding energy: stronger, weaker, or no_change relative to the reference.
qc_pass
True when all required sequence/matrix values are finite and the row passed package QC.
Quality control
The paper’s stated sequencing QC was applied before matrix inference: per-base PHRED score >20, with approximately 300,000–500,000 useful reads per bin after filtering. The paper used four FACS bins (each approximately 15% of the fluorescence histogram), at least 10^6 transformants per library, and Bayesian energy-matrix inference with three non-overlapping split replicates. Package QC retained only rows with a 21-bp sequence, valid A/C/G/T matrix entries, finite replicate aggregates, and qc_pass=true. Published validation records are retained only when their reported prediction/measurement values are finite.
Curation notes
The published S1 Models archive is the source for matrix rows. Matrix values are converted to alternative-base minus reference-base values at each position. Negative values indicate a more favorable/more negative predicted LacI binding energy. The first experiment includes published flow-cytometry validation records as record_type=published_flow_cytometry_validation; these are not raw Sort-Seq sequence counts and are kept distinct from MPRA matrix rows. R=130 full-promoter matrices are the primary kBT-scaled results; other model contexts are reported in au as supplied by the authors.