Upload DNA sequence
Upload the nucleotide file in FASTA format (EMBL and GenBank also supported). Parsed securely — your data stays private.
Upload a DNA sequence — see how each codon maps to a secondary structure type. Alpha-helix, beta-strand, pi-helix, 310-helix — mapped directly from codons. A research tool for studying the codon–SS relationship.
No installation needed. Works entirely in the browser.
Upload the nucleotide file in FASTA format (EMBL and GenBank also supported). Parsed securely — your data stays private.
Each DNA codon maps to one of 4 secondary structure types via the DNE_GROUPS lookup table — alpha-helix, beta-strand, pi-helix, 310-helix. Deterministic and instant.
View the 2D secondary structure scheme, composition statistics and detailed assignment table. Export results as CSV.
View the SS color scale for each residue, composition statistics, and a detailed distribution table.
All your predictions are saved. Revisit, re-download or compare results any time.
Files stored by UUID, JWT auth, bcrypt passwords. Your sequences never leave your account.
Color-coded amino acid grid with Code-4 and Code-9 modes — matches the compositional genetic code table.
Heavy computation runs server-side. Close the tab — your job keeps running.
Color-coded 2D scheme with Code-4 and Code-9 modes. See the deterministic codon-to-structure mapping at a glance.
The only tool that maps DNA codons to secondary structure types.
Study how synonymous codons — same amino acid, different DNA — map to different secondary structure types. Explore the relationship for any gene.
Study the relationship between codon usage patterns and protein secondary structure composition. Compare codon bias across organisms and its structural implications.
Investigate how synonymous SNPs — invisible to protein-based tools — show different codon–SS mappings. A hypothesis-generating signal for further validation.
Demonstrate how DNA codon sequence determines protein secondary structure. Instant, visual, interactive — no installation required.
Add a unique data point to your analysis pipeline. PicoFold's codon–SS mapping offers a signal that no protein-sequence tool provides.
The ring-granular model maps each DNA codon to a secondary structure type — the third nucleotide is the key determinant. This compositional genetic code has been validated against DSSP ground truth on 113 proteins (21,797 residues). Helix signal confirmed: G-ending codons show 39.5% helix vs C-ending 30.1%.
No other tool maps secondary structure from DNA — all competitors (AlphaFold, PSIPRED, JPred) work from protein sequence and achieve significantly higher accuracy (82–85% vs PicoFold's 37.7%). PicoFold is the only direct DNA-to-SS mapping tool. This means PicoFold can show how synonymous codon substitutions — same amino acid, different DNA — produce different SS assignments, a unique research capability.
for i in range(0, len(dna), 3):
b1, b2, b3 = dna[i:i+3]
helix = DNE_GROUPS[b1][b2][b3]
amino = DNE_AMI[b1][b2][b3]
if amino == "TKD": # stop codon
break
RotateAtoms(helix)
AddAtoms(ATOMS_COMMON)
AddAtoms(SIDE_CHAIN_MAP[b1*4+b2][b3])
PicoFold is a web-based research tool that maps DNA codons to secondary structure types — not from protein sequence, but directly from DNA. It is the only tool with this capability. All other secondary structure predictors (PSIPRED ~84%, JPred ~82%, NetSurfP ~85%) require protein amino acid sequence as input and achieve significantly higher accuracy. PicoFold takes DNA (FASTA, EMBL or GenBank format) and maps each codon to a SS type using the Ring-Granular Model.
This makes PicoFold uniquely relevant for codon–structure research, synonymous codon analysis, and molecular biology education — contexts where the specific DNA codon matters, not just the amino acid it encodes. Two synonymous codons encoding the same amino acid produce different SS assignments in PicoFold, because the model maps directly from the three-nucleotide codon to the local geometry.
PicoFold is developed by Nanoworld Laboratory. It is deterministic, instant (under 10ms), and requires no GPU. Validated on 113 proteins against DSSP ground truth (Q3=0.377, +3.1pp above random baseline). Helix signal confirmed; beta prediction is under development. Current focus: secondary structure analysis. 3D visualization is experimental.
Do synonymous codons affect protein structure? Evidence says yes — and PicoFold is the only tool for exploring this question. Chi-square analysis shows 6 of 18 amino acid families have significant codon–SS associations. Nature Communications 2022 confirms 57/87 synonymous codon pairs show different backbone angles. Learn more about the validation →
Every new account gets 5 free prediction credits — no card required.
Create a free account and run your first analysis in under a minute.