Experiment / E05HWSS4YOther

Model-directed evolution validation of random and native 5′ UTRs

Deep learning of the regulatory grammar of yeast 5′ untranslated regions from 500,000 random sequences

Start, midpoint, and endpoint sequences selected by the CNN from random and native 5′ UTRs were synthesized in the same HIS3 reporter system and tested under three 3-AT selection strengths. The processed table contains the complete-value rows from the supplemental forward-engineering growth table and derives within-parent changes from the starting sequence.

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

SD-His-Leu growth selection with 1.5, 3, or 5 mM 3-AT

The authors iteratively selected single-nucleotide substitutions with the CNN, then experimentally tested start, midpoint, and endpoint variants. Supplemental Table S4 reports log2 growth values at 1.5, 3, and 5 mM 3-AT; the processed table adds the mean across conditions and within-parent differences versus the Start row. The GEO Evolved_library files are model-evolution trajectories with predicted growth and Round0_Measured_Growth, not independent observed per-variant count tables, and are retained separately in raw_data.

Arrayed validation of model-designed episomal p415-CYC1-HIS3 reporters: computationally evolved random or native 5′ UTR sequences were individually synthesized, cloned immediately upstream of HIS3, transformed into BY4741 yeast, and assayed by competitive growth/sequence enrichment under multiple 3-AT concentrations.

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
element_id
Stable package identifier formed by prefixing the source UTR name with its library type.
source_utr_name
UTR name reported in the forward-engineering supplemental table, including its sequence-stage suffix.
parent_id
Starting UTR identifier obtained by removing the final underscore-delimited sequence-stage suffix from source_utr_name.
library_type
Source library of the parent sequence: native or random.
evolution_stage
Source prediction stage of the tested sequence: Start, Mid, or End.
source_evolution_step
Numeric suffix in source_utr_name, retained as the source sequence/evolution step label.
sequence
The tested forward-engineered 5′ UTR sequence.
sequence_length
Length of the tested UTR sequence in nucleotides.
log2_growth_1p5mM_3AT
Reported log2 growth/enrichment measured under 1.5 mM 3-AT selection.
log2_growth_3mM_3AT
Reported log2 growth/enrichment measured under 3 mM 3-AT selection.
log2_growth_5mM_3AT
Reported log2 growth/enrichment measured under 5 mM 3-AT selection.
mean_log2_growth
Arithmetic mean of the three reported log2 growth values across 1.5, 3, and 5 mM 3-AT.
delta_log2_growth_1p5mM_vs_start
Growth at 1.5 mM 3-AT minus the matching parent’s Start growth; blank if no complete Start baseline was retained.
delta_log2_growth_3mM_vs_start
Growth at 3 mM 3-AT minus the matching parent’s Start growth; blank if no complete Start baseline was retained.
delta_log2_growth_5mM_vs_start
Growth at 5 mM 3-AT minus the matching parent’s Start growth; blank if no complete Start baseline was retained.

Quality control

The source supplemental PDF contains 574 candidate rows. Four rows were excluded because one or more reported growth values were blank or #N/A: the YNL334C:47:0 start and midpoint rows, the YPL119C-A:50:0 endpoint row missing its 5 mM value, and random endpoint 206_31 with #N/A at 5 mM. The 570 retained rows have unique source identifiers, A/C/G/T sequences of length 4–50 nt, and complete finite growth values; they comprise 294 random-library rows and 276 native-library rows. The derived delta columns are left blank where a retained row has no complete Start baseline.

Curation notes

The downloaded supplementary file is named Supplemental_Table_S4 but its internal heading says Supplemental Table 3. Four incomplete/#N/A rows were removed at row level; the complete end row for native parent YNL334C:47:0 is retained but has blank deltas because its Start row was incomplete. This table reports measured growth for the experimentally validated subset, whereas the full model-evolution paths remain in the two GEO GSM2793756 CSV.gz files.

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.