v1.7 · Program 04C · Browser-first · Design–Build–Test–Learn

RNA DBTL & Structure Intelligence

Register construct versions, inspect RNA folding hypotheses and imported protein models, link measured experiments and QC evidence, and learn from the recorded data without inventing predictive performance.

Construct version historySeqFold local RNA pairingDot-bracket importAlphaFold DB linksLocal PDB confidence reviewProject Hub

DBTL program workflow

1 · DESIGNCreate and version an RNA transcript hypothesis
2 · BUILDKeep immutable sequence snapshots and source references
3 · TESTMatch measurements and QC by sample and construct IDs
4 · LEARNCompare observed outcomes and gaps; plan the next iteration
Local browser processing only. This release does not run AlphaFold inference, ViennaRNA thermodynamic folding, train ML models, or identify the “best RNA.” AlphaFold database records require sequence/protein identity review. Do not enter patient-identifiable or confidential clinical data.
0Construct versions
0Measured results
0Structure assessments

Design · immutable snapshots

1. Register RNA construct

Import the last design from UTR Studio or paste an unambiguous RNA/DNA sequence. Each save creates a new version; existing versions are never overwritten.

Construct versionParentLengthFingerprint*

*Fingerprint is for accidental-change detection only; it is not cryptographic authentication. A saved design is a research record, not a GMP electronic signature.

Structure · RNA and encoded protein

2. Structure intelligence

Open official RNAfold ↗

This computes the SeqFold/Nussinov maximum-base-pair heuristic, not folding free energy or ViennaRNA RNAfold. A 350-nt window is not a full transcript.

No RNA structure assessed.

Import RNAfold / RNAstructure dot-bracket

Paste an externally predicted dot-bracket line with optional parenthesized energy. Set the RNA window coordinates above to match the imported structure. External results are not sent to NEXUS.

AlphaFold DB structure review

Search AlphaFold DB ↗

AlphaFold DB is a separately hosted protein-model resource. This tool does not submit a sequence for inference or automatically establish a match to the encoded protein.

No PDB analyzed.
Test · observations and traceability

3. Record a measured result

Associate each observation with a registered construct, a unique sample ID, units, and the original evidence reference. This is not an instrument raw-data archive.

Open QC Analytics ↗Import instrument results ↗
ExperimentVersionSampleEndpointResultMatched QC records
Learn · descriptive evidence

4. Compare available evidence

Summaries are grouped only by identical construct version, assay label and units. They do not rank constructs or infer biological significance.

VersionAssaynMeanSample SDUnit
Comparisons depend on assay normalization, independent biological replicates, controls and study design. A higher measured signal is not necessarily a better RNA. Do not treat multiple technical wells as independent biological replicates.
Open connected Project Hub ↗

Research and attribution resources

ViennaRNA RNAfold documentation ↗RNAstructure ↗AlphaFold Protein Structure Database ↗NEXUS CDS & UTR Designer ↗
AlphaFold DB predictions: DeepMind Technologies / EMBL-EBI, CC BY 4.0 attribution required. Protein model confidence (pLDDT) cannot establish mRNA stability, expression, antigenicity or vaccine efficacy. External sequence submission and licensed predictive inference require a future backend and separate governance review.