| name | bio-moldesign |
| description | Design and QC molecular-biology reagents with Seqeron's MolTools server (47 tools) — PCR primer pairs and hybridization probes, CRISPR guide RNAs with off-target and specificity scoring, codon optimization of a CDS for a target organism (CAI/GC before-after), restriction-site finding and digest simulation, and melting-temperature (Tm) with salt correction. Use for "design primers for this template", "find/design CRISPR guides for this target and check off-targets", "codon-optimize this gene for E. coli / yeast / human", "where does EcoRI cut / simulate a digest", "design a qPCR/FISH probe", "compute the salt-corrected Tm", primer-dimer / hairpin / 3'-stability QC, and choosing compatible restriction enzymes. Dual-mode: MCP tool calls + equivalent Seqeron.Genomics C# Method IDs. |
| allowed-tools | Read, Grep, Glob, Bash |
bio-moldesign — molecular design & QC (MolTools)
Routing + orchestration layer over the MolTools MCP server (47 tools) and the
Seqeron.Genomics.MolTools C# namespace. It gives dual-mode recipes; it does not
restate schemas — read the per-tool doc for exact I/O, and delegate all rigor to bio-rigor.
- Compute with tools, never by eye. Parse/validate/compute via a tool or
Method ID. See
bio-rigor for the six rules (tool-only, envelope, provenance, cross-check, units, disclaimer).
- Don't guess tool names. Confirm with
python3 scripts/skills/find-tool.py <kw> --server moltools
or via seqeron-discovery.
- Coordinates in MolTools tools are 0-based (target/region start/end); Tm is °C, GC is % (or fraction in codon tools).
Envelope / STOP rule
One MolTools unit is guarded: PROBE-DESIGN-001 (probe design), MinimumMode Moderate
(default). It is correct-but-incomplete: it does not model quantitative MGB ΔTm or
dual-quencher labelling (empirical/proprietary, no closed form). If a task needs a real MGB /
dual-quencher probe Tm, STOP and report the limitation — do not force a number. See
docs/Validation/LIMITATIONS.md (row PROBE-DESIGN-001)
and bio-rigor reference/envelope.md.
Decision guide — task → tool family
| You need to… | Family | Entry tool(s) |
|---|
| Design a PCR primer pair for a target region | primer | design_primers → QC below |
| Enumerate / score candidate primers, or QC one primer | primer | generate_primer_candidates, evaluate_primer |
| Primer QC: dimer / hairpin / 3′ stability / repeats | primer | primer_dimer, hairpin_potential, three_prime_stability, longest_homopolymer, longest_dinucleotide_repeat |
| Melting temperature (basic / salt-corrected) | primer | primer_melting_temperature, primer_melting_temperature_salt |
| Design CRISPR guide RNAs for a region | crispr | design_guide_rnas → find_off_targets → crispr_specificity_score |
| Find PAM sites / evaluate one guide / system info | crispr | find_pam_sites, evaluate_guide_rna, crispr_system_info |
| Codon-optimize a CDS for an organism | codon | optimize_codons (reports CAI+GC before/after) |
| Codon usage analysis (CAI, RSCU, ENC, rare codons) | codon | codon_adaptation_index, rscu, effective_number_of_codons, find_rare_codons |
| Restriction sites / digest / enzyme choice | restriction | find_restriction_sites, restriction_digest, digest_summary, compatible_enzymes |
| Design / validate a hybridization probe | probe | design_probes, validate_probe, analyze_oligo (see STOP rule for MGB) |
Full per-tool map: reference/tool-map.md. Fuller recipes + parameter
guidance: reference/pipelines.md.
Canonical dual-mode pipelines
Each pipeline is ordered; MCP tool = MolTools server, C# = Seqeron.Genomics.MolTools.
(a) PCR primer pair design + QC
Goal: a valid, well-behaved primer pair flanking a target, with a Tm-difference report.
- [MCP]
design_primers(template, target_start, target_end) → for each of fwd/rev:
primer_melting_temperature_salt, primer_dimer(fwd,rev), hairpin_potential, three_prime_stability.
- [C# API]
PrimerDesigner.DesignPrimers → PrimerDesigner.CalculateMeltingTemperatureWithSalt,
PrimerDesigner.HasPrimerDimer, PrimerDesigner.HasHairpinPotential, PrimerDesigner.Calculate3PrimeStability.
- Key params: 0-based inclusive
target_start/target_end; defaults 18–25 nt, 40–60% GC, 57–63 °C;
pair "compatible" when ΔTm ≤ 5 °C and no 3′ dimer. Report |Tm_fwd − Tm_rev| (°C, 1 dp).
- Provenance: design_primers → primer_melting_temperature_salt(×2) → primer_dimer → hairpin_potential, with all params.
(b) CRISPR guide design → off-targets → specificity → rank
Goal: rank guide RNAs targeting a region by on-target score and off-target specificity.
- [MCP]
design_guide_rnas(sequence, region_start, region_end, system_type) → for each guide:
find_off_targets(guide, genome, max_mismatches) → crispr_specificity_score(guide, genome) → rank by (score, specificity).
- [C# API]
CrisprDesigner.DesignGuideRnas → CrisprDesigner.FindOffTargets → CrisprDesigner.CalculateSpecificityScore.
- Key params: 0-based inclusive region; default system
SpCas9; max_mismatches 0..5 (default 3);
specificity uses maxMismatches=4, seed = PAM-proximal 12 bp. Keep genome ≲ 1 Mb (O(genome×guide)).
- Provenance: design_guide_rnas → find_off_targets → crispr_specificity_score per guide, + system + params.
(c) Codon-optimize a CDS for a target organism
Goal: raise CAI for the host while staying in a GC band; report CAI/GC before→after.
- [MCP]
optimize_codons(coding_sequence, target_organism, strategy, gc_target_min/max) → returns
original/optimized CAI + GC + changed codons; optionally find_rare_codons first to inspect.
- [C# API]
CodonOptimizer.OptimizeSequence; standalone CAI via CodonUsageAnalyzer.CalculateCai
or CodonOptimizer.CalculateCAI (organism table).
- Key params: presets
EColiK12/Yeast/Human (or custom RNA table); strategy default BalancedOptimization
(MaximizeCAI deterministic; HarmonizeExpression non-deterministic — set seed/repeat); GC band default 0.40–0.60.
- Provenance: optimize_codons(preset, strategy, GC band) → report ΔCAI, ΔGC, changedCodons.
(d) Restriction analysis: find sites / simulate digest
Goal: locate cut sites for chosen enzymes and produce fragment list.
- [MCP]
find_restriction_sites(sequence, enzyme_names) → restriction_digest(sequence, enzyme_names)
→ digest_summary; choose enzymes with compatible_enzymes / get_enzyme.
- [C# API]
RestrictionAnalyzer.FindSites → RestrictionAnalyzer.Digest → RestrictionAnalyzer.GetDigestSummary.
- Key params: linear molecule;
k distinct forward cuts → k+1 fragments. Site positions are 0-based.
- Provenance: find_restriction_sites(enzymes) → restriction_digest(enzymes) → fragment table.
(e) Probe design + salt-corrected Tm
Goal: rank hybridization probes for a target and report a defensible Tm.
- [MCP]
design_probes(target_sequence, parameters=preset, max_probes) → validate_probe /
analyze_oligo per candidate → primer_melting_temperature_salt for a salt-aware Tm.
- [C# API]
ProbeDesigner.DesignProbes → ProbeDesigner.ValidateProbe / ProbeDesigner.AnalyzeOligo
→ PrimerDesigner.CalculateMeltingTemperatureWithSalt.
- Key params: presets
Microarray(default)/FISH/NorthernBlot/qPCR/SouthernBlot; salt Tm adds
16.6·log10([Na+]/1000). STOP if the assay needs MGB ΔTm / dual-quencher (PROBE-DESIGN-001 envelope).
- Provenance: design_probes(preset) → validate_probe → primer_melting_temperature_salt([Na+]).
End-to-end grounded example (extends the README primer-QC workflow)
The README shows validate-DNA → GC% → Tm → ΔTm for a given primer pair
(docs/mcp/README.md §"PCR primer QC"). Extend it into a full
design + QC task: "Design a primer pair flanking target [100,150) in this template and QC it."
Tool / Method-ID chain (MCP → C#):
design_primers(template, 100, 150) → PrimerDesigner.DesignPrimers — yields best fwd + rev, isValid, productSize.primer_melting_temperature_salt(fwd) and (rev) → PrimerDesigner.CalculateMeltingTemperatureWithSalt — Tm at chosen [Na+].primer_dimer(fwd, rev) → PrimerDesigner.HasPrimerDimer — cross-dimer check.hairpin_potential(fwd) and (rev) → PrimerDesigner.HasHairpinPotential.
Expected-shape output (Markdown table + a compatibility line):
| id | length | gc_percent | tm_c | dimer | hairpin |
|---|---:|---:|---:|:---:|:---:|
| FWD | 25 | 52.00 | 59.3 | no | no |
| REV | 25 | 52.00 | 59.3 | no | no |
product_size_bp = 180 ; tm_diff_c = 0.0 ; pair_compatible = true
Then a Provenance block listing tools/Method-IDs in call order with params, plus the
alpha / not-for-clinical-use caveat from bio-rigor if the assay is decision-relevant.
Reference
