See how your RNA folds and binds

ProteinIQ supports these distinct evidence layers without treating them as interchangeable. You can run matched RNA methods, compare structures and interaction signals, and retain the energies, confidence values, coordinates, plots, tables, and files returned by each analysis. Start with an RNA sequence and the biological question you need to answer, then choose the relevant folding, modeling, interaction, design, or screening workflow. ProteinIQ keeps every result connected to the original sequence and ready for inspection, export, or experimental planning.

ProteinIQ supports RNA secondary-structure prediction, ensemble and accessibility analysis, RNA three-dimensional structure prediction, RNA–RNA interaction prediction, RNA inverse folding, and RNA-targeted small-molecule screening. Each method returns its own scores, structures, tables, and files rather than collapsing unlike evidence into one result.

Most RNA workflows start from RNA sequence in FASTA or plain-text form. Depending on the method, you may also provide a second RNA sequence, a dot-bracket structure, an alignment, supported folding constraints, known molecular partners, or named compounds with SMILES strings.

Yes. RNAfold, RNAsubopt, RNAplfold, and related ViennaRNA methods return base-pairing and ensemble evidence, while Chai-1, Boltz-2, OpenFold 3, and Protenix generate atomic-coordinate hypotheses. These outputs answer different questions and should be reviewed separately.

Yes. ProteinIQ can compare RNAcofold, RNAduplex, RNAup, and RNAplex outputs for two supplied RNA sequences. Results can include joint structures, duplex coordinates, accessibility terms, and interaction energies, depending on the selected method.

Yes. SMRTnet accepts named RNA sequences with dot-bracket structures and named compounds with SMILES strings, then returns binding probabilities, interaction calls, and the selected RNA window. These are prioritization signals, not measurements of binding affinity or activity.

Treat a predicted fold or coordinate model as a testable structural hypothesis. Review alternative folds, ensemble evidence, confidence, sequence context, known modifications, molecular partners, and agreement with probing, binding, cryo-EM, NMR, crystallography, or other experimental evidence when available.

Yes. Depending on the tool, ProteinIQ can export FASTA, dot-bracket structures, CSV or JSON tables, plots, CIF or PDB coordinates, confidence data, model files, and logs while preserving the identifiers and settings used for the run.

No. Computational RNA models support prioritization and experiment design. Structural probing, binding assays, activity measurements, cell-based studies, and high-resolution structural methods remain necessary before making biological or therapeutic claims.