Antibody Engineering & Optimization
AI-guided antibody optimization pipeline from preclinical lead to clinical candidate. Covers sequence humanization, structure modeling, affinity optimization, developability assessment, immunogenicity prediction, and manufacturing feasibility.
KEY PRINCIPLES:
- Report-first approach - Create optimization report before analysis
- Evidence-graded humanization - Score based on germline alignment and framework retention
- Developability-focused - Assess aggregation, stability, PTMs, immunogenicity
- Structure-guided - Use AlphaFold/PDB structures for CDR analysis
- Clinical precedent - Reference approved antibodies for validation
- Quantitative scoring - Developability score (0-100) combining multiple factors
- English-first queries - Always use English terms in tool calls, even if user writes in another language. Respond in user's language
LOOK UP, DON'T GUESS
When uncertain about any scientific fact, SEARCH databases first (PubMed, UniProt, ChEMBL, ClinVar, etc.) rather than reasoning from memory. A database-verified answer is always more reliable than a guess.
When to Use
Apply when user asks:
- "Humanize this mouse antibody sequence"
- "Optimize antibody affinity for [target]"
- "Assess developability of this antibody"
- "Predict immunogenicity risk for [sequence]"
- "Engineer bispecific antibody against [targets]"
- "Reduce aggregation in antibody formulation"
- "Design pH-dependent binding antibody"
- "Analyze CDR sequences and suggest mutations"
Critical Workflow Requirements
1. Report-First Approach (MANDATORY)
- Create the report file FIRST:
antibody_optimization_report.md
- Progressively update as analysis completes
- Output separate files:
optimized_sequences.fasta - All optimized variants
humanization_comparison.csv - Before/after comparison
developability_assessment.csv - Detailed scores
See REPORT_TEMPLATE.md for the full report template with section formats.
2. Documentation Standards (MANDATORY)
Every optimization MUST include per-variant documentation with:
- Original and optimized sequences
- Humanization score (% human framework)
- CDR preservation confirmation
- Metrics table (humanness, aggregation risk, predicted KD, immunogenicity)
- Data source citations
Phase 0: Tool Verification
Required Tools
| Tool |
Purpose |
Category |
IMGT_search_genes |
Germline gene identification |
Humanization |
IMGT_get_sequence |
Human framework sequences |
Humanization |
SAbDab_search_structures |
Antibody structure precedents |
Structure |
TheraSAbDab_search_by_target |
Clinical antibody benchmarks |
Validation |
ESMFold_predict_structure |
Fv structure modeling (VH-linker-VL scFv) |
Structure |
alphafold_get_prediction |
Retrieve a precomputed AlphaFold DB model by UniProt accession (target antigen) |
Structure |
iedb_search_epitopes |
Epitope identification |
Immunogenicity |
iedb_search_bcell |
B-cell epitope prediction |
Immunogenicity |
UniProt_get_entry_by_accession |
Target antigen information |
Target |
STRING_get_interaction_partners |
Protein interaction network |
Bispecifics |
PubMed_search_articles |
Literature precedents |
Validation |
CRITICAL: SOAP tools (IMGT, SAbDab, TheraSAbDab) require an operation parameter. See QUICK_START.md for correct usage.
Workflow Overview
Phase 1: Input Analysis & Characterization
├── Sequence annotation (CDRs, framework)
├── Species identification
├── Target antigen identification
├── Clinical precedent search
└── OUTPUT: Input characterization
↓
Phase 2: Humanization Strategy
├── Germline gene alignment (IMGT)
├── Framework selection
├── CDR grafting design
├── Backmutation identification
└── OUTPUT: Humanization plan
↓
Phase 3: Structure Modeling & Analysis
├── AlphaFold prediction
├── CDR conformation analysis
├── Epitope mapping
├── Interface analysis
└── OUTPUT: Structural assessment
↓
Phase 4: Affinity Optimization
├── In silico mutation screening
├── CDR optimization strategies
├── Interface improvement
└── OUTPUT: Affinity variants
↓
Phase 5: Developability Assessment
├── Aggregation propensity
├── PTM site identification
├── Stability prediction
├── Expression prediction
└── OUTPUT: Developability score
↓
Phase 6: Immunogenicity Prediction
├── MHC-II epitope prediction (IEDB)
├── T-cell epitope risk
├── Aggregation-related immunogenicity
└── OUTPUT: Immunogenicity risk score
↓
Phase 7: Manufacturing Feasibility
├── Expression level prediction
├── Purification considerations
├── Formulation stability
└── OUTPUT: Manufacturing assessment
↓
Phase 8: Final Report & Recommendations
├── Ranked variant list
├── Experimental validation plan
├── Next steps
└── OUTPUT: Comprehensive report
Phase 1: Input Analysis & Characterization
Goal: Annotate sequences, identify species/germline, find clinical precedents.
Key steps:
- Annotate CDRs using an antibody-numbering tool and map labels back to sequence indices (IMGT CDR-H1: 27-38, CDR-H2: 56-65, CDR-H3: 105-117). These labels are not literal array offsets, especially with insertions or terminal extensions.
- Identify closest human germline genes via
IMGT_search_genes
- Search clinical precedents via
TheraSAbDab_search_by_target
- Get target antigen info via
UniProt_get_entry_by_accession
Output: Sequence information table, CDR annotation, target info, clinical precedent list.
See WORKFLOW_DETAILS.md Phase 1 for code examples.
Phase 2: Humanization Strategy
Goal: Select human framework, design CDR grafting, identify backmutations.
Key steps:
- Search IMGT for IGHV/IGKV human germline genes
- Score candidate frameworks by identity, CDR compatibility, and clinical use
- Design CDR grafting onto selected framework
- Identify Vernier zone residues that may need backmutation (positions 2, 27-30, 47-48, 67, 69, 71, 78, 93-94)
- Generate at least 2 variants: full humanization and with key backmutations
- Calculate humanization score (framework humanness, CDR preservation, T-cell epitopes, aggregation risk)
Output: Framework selection rationale, grafting design, backmutation analysis, humanized sequences.
See WORKFLOW_DETAILS.md Phase 2 for code examples.
Phase 3: Structure Modeling & Analysis
Goal: Predict structure, analyze CDR conformations, map epitope.
Key steps:
- Predict Fv structure via
ESMFold_predict_structure (VH-linker-VL scFv; alphafold_get_prediction only retrieves precomputed models by UniProt accession)
- Assess pLDDT scores by region (framework, CDRs, interface)
- Classify CDR canonical structures and calculate RMSD
- Search known epitopes via
iedb_search_epitopes
- Compare with clinical antibody structures via
SAbDab_search_structures
Output: Structure quality table, CDR conformation analysis, epitope mapping, structural comparison.
See WORKFLOW_DETAILS.md Phase 3 for code examples.
Phase 4: Affinity Optimization
Goal: Design affinity-improving mutations via computational screening.
Key steps:
- Identify interface residues (distance cutoff 4.5 A)
- Screen all amino acid substitutions at CDR interface positions
- Rank by computed binding-energy change only when the method and calibration support it; state the sign convention and uncertainty. An illustrative cutoff is not a universal binding gate.
- Design combination strategy: single -> double -> triple mutants
- Consider CDR-H3 extension, tyrosine enrichment, salt bridge formation
- Optional: specify stronger acidic binding or neutral binding with acidic release before considering histidine substitutions. Neither direction follows automatically from His pKa. Validate complete bound/free protonation assumptions with controls and matched assays; use Protein Design Campaign and its pH reference.
Output: Ranked mutation hypotheses, combination strategy, computed metrics where supported, and explicit unverified affinity/conditional-binding claims. Do not invent KD ranges or fold improvements.
See WORKFLOW_DETAILS.md Phase 4 for code examples.
Phase 5: Developability Assessment
Goal: Comprehensive developability scoring (0-100) across five dimensions.
Key steps:
- Aggregation, pI, hydrophobic patches (sequence-computable): run
scripts/developability.py --seq <VH/VL>. It returns AGGRESCAN aggregation-prone regions (real per-residue propensity), isoelectric point, and Kyte-Doolittle hydrophobic patches.
- PTM liability: Scan for deamidation (NG/NS), isomerization (DG/DS), oxidation (Met/Trp), N-glycosylation (N-X-S/T)
- Stability (Tm) and 4. Expression titer: NOT computable from sequence alone — use external predictors (sequence→Tm/titer ML tools; report only if actually run). Do not fabricate Tm/titer numbers.
- Solubility: external solubility predictor (e.g. NetSolP)
Scoring: Weighted average (aggregation 0.30, PTM 0.25, stability 0.20, expression 0.15, solubility 0.10).
Tiers: T1 (>75), T2 (60-75), T3 (<60).
Output: Component scores, overall score, tier classification, mitigation recommendations.
See WORKFLOW_DETAILS.md Phase 5 and CHECKLISTS.md for scoring details.
Phase 6: Immunogenicity Prediction
Goal: Predict immunogenicity risk and design deimmunization strategy.
Key steps:
- Scan 9-mer peptides against IEDB for MHC-II binding epitopes
- Count non-human residues in framework regions
- Assess aggregation-related immunogenicity
- Calculate total risk score (0-100, lower is better): Low <30, Medium 30-60, High >60
- Propose deimmunization mutations (remove T-cell epitopes while preserving CDRs)
- Compare with clinical precedent ADA rates
Output: T-cell epitope list, risk score breakdown, deimmunization strategy, clinical comparison.
See WORKFLOW_DETAILS.md Phase 6 for code examples.
Phase 7: Manufacturing Feasibility
Goal: Assess expression, purification, formulation, and CMC feasibility.
Key steps:
- Assess codon optimization for CHO, identify rare codons --
CodonUsage_get_optimal_codons(taxid=<Cricetulus griseus NCBI taxid>) gives the CHO-preferred codon per amino acid to compare against; resolve the taxid first if not already known rather than assuming one. CodonUsage_get_table (same taxid) additionally gives per-codon usage frequency, needed to flag which codons in the actual sequence are rare in CHO specifically -- human codon-usage tables (e.g. from tooluniverse-sequence-analysis) are the wrong reference here, rare-in-human is not the same set as rare-in-CHO
- Design signal peptide
- Plan 3-step purification: Protein A capture -> cation exchange polishing -> viral nanofiltration
- Recommend formulation (buffer, pH, stabilizer, tonicity)
- Define analytical characterization panel (SEC-MALS, CEX, CE-SDS, SPR, DSF)
- Estimate CMC timeline and costs (typically 18-24 months, $1.65-2.65M to IND)
Output: Expression assessment, purification strategy, formulation recommendation, CMC timeline.
See MANUFACTURING.md for detailed manufacturing content and WORKFLOW_DETAILS.md Phase 7 for code.
Phase 8: Final Report & Recommendations
Goal: Compile all findings into a ranked recommendation with validation plan.
Key outputs:
- Top candidate with key metrics (humanness, affinity, developability, immunogenicity, stability, expression)
- Key improvements table comparing original vs. optimized
- Experimental validation plan: In vitro (3-4 months) -> Lead optimization (2-3 months) -> Preclinical (6-12 months)
- Backup variants with profiles and recommendations
- IP considerations: FTO analysis, CDR novelty, patentability
- Next steps: Immediate (month 1-3), short-term (4-6), long-term (7-24)
See REPORT_TEMPLATE.md for the full report template.
Tool Reference
IMGT Tools
IMGT_search_genes: Search germline genes (IGHV, IGKV, etc.)
IMGT_get_sequence: Get germline sequences
IMGT_get_gene_info: Database information
Antibody Databases
SAbDab_search_structures: Search antibody structures
SAbDab_get_structure: Get structure details
TheraSAbDab_search_therapeutics: Search by name
TheraSAbDab_search_by_target: Search by target antigen
Immunogenicity
iedb_search_epitopes: Search epitopes
iedb_search_bcell: B-cell epitopes
iedb_search_mhc: MHC-II epitopes
iedb_get_epitope_references: Citations
Structure & Target
ESMFold_predict_structure: Fv structure prediction from a sequence
alphafold_get_prediction: Retrieve a precomputed AlphaFold DB model by UniProt accession
UniProt_get_entry_by_accession: Target info
RCSBData_get_entry: Experimental structures
Reagent Verification (Antibody Registry)
AntibodyRegistry_search: Find validated commercial antibodies for a target (by target gene/protein name, antibody name, or vendor keyword). Returns each hit's RRID, target (with UniProt/Entrez id), clonality, host organism, applications, vendor/catalog number, and defining citation. Useful during Phase 1 (target/precedent research) to see what antibody reagents already exist against a target antigen before designing a new one, or during Phase 8 (validation planning) to identify a benchmark reagent.
AntibodyRegistry_get_by_rrid: Resolve a specific RRID (e.g. AB_2298772, accepts the bare numeric id or RRID:AB_... form too) cited in a methods section to its full record — verify which exact antibody a paper used rather than assuming from the name alone. Live-verified: both tools return real records (e.g. AntibodyRegistry_search(query="GFAP") -> a real GFAP rabbit mAb with UniProt target P14136).
SciCrunch_resolve_rrid: General RRID resolver for everything that ISN'T an antibody — cell lines (CVCL_ prefix), model organisms, software/tools (SCR_ prefix), and plasmids. Use it when a clinical-precedent or validation paper's methods section also cites the cell line or software used alongside an antibody. It will still resolve an AB_ RRID, but only returns generic fields (name, description, categories, proper_citation) — no target/clonality/host — so prefer AntibodyRegistry_get_by_rrid for antibody RRIDs specifically. Live-verified: SciCrunch_resolve_rrid(rrid="CVCL_0045") returns the real HEK293 cell-line record (12 synonyms, proper_citation: "NCBI_Iran Cat# C497, RRID:CVCL_0045"); on the same AB_2298772 antibody RRID above it returns only name/description/categories with no target info, confirming the richer-antibody-fields distinction.
Systems Biology (for Bispecifics)
STRING_get_interaction_partners: Protein interactions
STRING_get_enrichment: Pathway analysis
Reference Files
| File |
Contents |
QUICK_START.md |
Getting started guide, SOAP tool parameters, Python SDK and MCP usage |
WORKFLOW_DETAILS.md |
Code examples for all 8 phases |
REPORT_TEMPLATE.md |
Full report template with section formats and example tables |
MANUFACTURING.md |
Detailed manufacturing content (expression, purification, formulation, CMC) |
EXAMPLES.md |
Complete clinical scenario examples (humanization, affinity, bispecific) |
CHECKLISTS.md |
Evidence grading, completeness checklists, scoring details, special considerations |
1---23name: tooluniverse-antibody-engineering-24description: "Therapeutic antibody engineering and optimization, lead-to-clinical-candidate. Covers sequence humanization (germline alignment, framework retention), affinity maturation, developability (aggregation, stability, PTMs), structure modeling (AlphaFold/PDB CDR analysis), immunogenicity prediction, and manufacturing feasibility. Use for biologic-drug optimization, mAb design review, biosimilar engineering, and clinical-precedent comparison."5metadata:6 source: https://skillmd.com/skills/mims-harvard/tooluniverse-antibody-engineering-27---89# Antibody Engineering & Optimization1011AI-guided antibody optimization pipeline from preclinical lead to clinical candidate. Covers sequence humanization, structure modeling, affinity optimization, developability assessment, immunogenicity prediction, and manufacturing feasibility.1213**KEY PRINCIPLES**:141. **Report-first approach** - Create optimization report before analysis152. **Evidence-graded humanization** - Score based on germline alignment and framework retention163. **Developability-focused** - Assess aggregation, stability, PTMs, immunogenicity174. **Structure-guided** - Use AlphaFold/PDB structures for CDR analysis185. **Clinical precedent** - Reference approved antibodies for validation196. **Quantitative scoring** - Developability score (0-100) combining multiple factors207. **English-first queries** - Always use English terms in tool calls, even if user writes in another language. Respond in user's language2122---2324## LOOK UP, DON'T GUESS25When uncertain about any scientific fact, SEARCH databases first (PubMed, UniProt, ChEMBL, ClinVar, etc.) rather than reasoning from memory. A database-verified answer is always more reliable than a guess.2627---2829## When to Use3031Apply when user asks:32- "Humanize this mouse antibody sequence"33- "Optimize antibody affinity for [target]"34- "Assess developability of this antibody"35- "Predict immunogenicity risk for [sequence]"36- "Engineer bispecific antibody against [targets]"37- "Reduce aggregation in antibody formulation"38- "Design pH-dependent binding antibody"39- "Analyze CDR sequences and suggest mutations"4041---4243## Critical Workflow Requirements4445### 1. Report-First Approach (MANDATORY)46471. **Create the report file FIRST**: `antibody_optimization_report.md`482. **Progressively update** as analysis completes493. **Output separate files**:50 - `optimized_sequences.fasta` - All optimized variants51 - `humanization_comparison.csv` - Before/after comparison52 - `developability_assessment.csv` - Detailed scores5354See `REPORT_TEMPLATE.md` for the full report template with section formats.5556### 2. Documentation Standards (MANDATORY)5758Every optimization MUST include per-variant documentation with:59- Original and optimized sequences60- Humanization score (% human framework)61- CDR preservation confirmation62- Metrics table (humanness, aggregation risk, predicted KD, immunogenicity)63- Data source citations6465---6667## Phase 0: Tool Verification6869### Required Tools7071| Tool | Purpose | Category |72|------|---------|----------|73| `IMGT_search_genes` | Germline gene identification | Humanization |74| `IMGT_get_sequence` | Human framework sequences | Humanization |75| `SAbDab_search_structures` | Antibody structure precedents | Structure |76| `TheraSAbDab_search_by_target` | Clinical antibody benchmarks | Validation |77| `ESMFold_predict_structure` | Fv structure modeling (VH-linker-VL scFv) | Structure |78| `alphafold_get_prediction` | Retrieve a precomputed AlphaFold DB model by UniProt accession (target antigen) | Structure |79| `iedb_search_epitopes` | Epitope identification | Immunogenicity |80| `iedb_search_bcell` | B-cell epitope prediction | Immunogenicity |81| `UniProt_get_entry_by_accession` | Target antigen information | Target |82| `STRING_get_interaction_partners` | Protein interaction network | Bispecifics |83| `PubMed_search_articles` | Literature precedents | Validation |8485**CRITICAL**: SOAP tools (IMGT, SAbDab, TheraSAbDab) require an `operation` parameter. See `QUICK_START.md` for correct usage.8687---8889## Workflow Overview9091```92Phase 1: Input Analysis & Characterization93├── Sequence annotation (CDRs, framework)94├── Species identification95├── Target antigen identification96├── Clinical precedent search97└── OUTPUT: Input characterization98 ↓99Phase 2: Humanization Strategy100├── Germline gene alignment (IMGT)101├── Framework selection102├── CDR grafting design103├── Backmutation identification104└── OUTPUT: Humanization plan105 ↓106Phase 3: Structure Modeling & Analysis107├── AlphaFold prediction108├── CDR conformation analysis109├── Epitope mapping110├── Interface analysis111└── OUTPUT: Structural assessment112 ↓113Phase 4: Affinity Optimization114├── In silico mutation screening115├── CDR optimization strategies116├── Interface improvement117└── OUTPUT: Affinity variants118 ↓119Phase 5: Developability Assessment120├── Aggregation propensity121├── PTM site identification122├── Stability prediction123├── Expression prediction124└── OUTPUT: Developability score125 ↓126Phase 6: Immunogenicity Prediction127├── MHC-II epitope prediction (IEDB)128├── T-cell epitope risk129├── Aggregation-related immunogenicity130└── OUTPUT: Immunogenicity risk score131 ↓132Phase 7: Manufacturing Feasibility133├── Expression level prediction134├── Purification considerations135├── Formulation stability136└── OUTPUT: Manufacturing assessment137 ↓138Phase 8: Final Report & Recommendations139├── Ranked variant list140├── Experimental validation plan141├── Next steps142└── OUTPUT: Comprehensive report143```144145---146147## Phase 1: Input Analysis & Characterization148149**Goal**: Annotate sequences, identify species/germline, find clinical precedents.150151**Key steps**:1521. Annotate CDRs using an antibody-numbering tool and map labels back to sequence indices (IMGT CDR-H1: 27-38, CDR-H2: 56-65, CDR-H3: 105-117). These labels are not literal array offsets, especially with insertions or terminal extensions.1532. Identify closest human germline genes via `IMGT_search_genes`1543. Search clinical precedents via `TheraSAbDab_search_by_target`1554. Get target antigen info via `UniProt_get_entry_by_accession`156157**Output**: Sequence information table, CDR annotation, target info, clinical precedent list.158159See `WORKFLOW_DETAILS.md` Phase 1 for code examples.160161---162163## Phase 2: Humanization Strategy164165**Goal**: Select human framework, design CDR grafting, identify backmutations.166167**Key steps**:1681. Search IMGT for IGHV/IGKV human germline genes1692. Score candidate frameworks by identity, CDR compatibility, and clinical use1703. Design CDR grafting onto selected framework1714. Identify Vernier zone residues that may need backmutation (positions 2, 27-30, 47-48, 67, 69, 71, 78, 93-94)1725. Generate at least 2 variants: full humanization and with key backmutations1736. Calculate humanization score (framework humanness, CDR preservation, T-cell epitopes, aggregation risk)174175**Output**: Framework selection rationale, grafting design, backmutation analysis, humanized sequences.176177See `WORKFLOW_DETAILS.md` Phase 2 for code examples.178179---180181## Phase 3: Structure Modeling & Analysis182183**Goal**: Predict structure, analyze CDR conformations, map epitope.184185**Key steps**:1861. Predict Fv structure via `ESMFold_predict_structure` (VH-linker-VL scFv; `alphafold_get_prediction` only retrieves precomputed models by UniProt accession)1872. Assess pLDDT scores by region (framework, CDRs, interface)1883. Classify CDR canonical structures and calculate RMSD1894. Search known epitopes via `iedb_search_epitopes`1905. Compare with clinical antibody structures via `SAbDab_search_structures`191192**Output**: Structure quality table, CDR conformation analysis, epitope mapping, structural comparison.193194See `WORKFLOW_DETAILS.md` Phase 3 for code examples.195196---197198## Phase 4: Affinity Optimization199200**Goal**: Design affinity-improving mutations via computational screening.201202**Key steps**:2031. Identify interface residues (distance cutoff 4.5 A)2042. Screen all amino acid substitutions at CDR interface positions2053. Rank by computed binding-energy change only when the method and calibration support it; state the sign convention and uncertainty. An illustrative cutoff is not a universal binding gate.2064. Design combination strategy: single -> double -> triple mutants2075. Consider CDR-H3 extension, tyrosine enrichment, salt bridge formation2086. Optional: specify stronger acidic binding or neutral binding with acidic release before considering histidine substitutions. Neither direction follows automatically from His pKa. Validate complete bound/free protonation assumptions with controls and matched assays; use [Protein Design Campaign](../tooluniverse-protein-design-campaign/SKILL.md) and its [pH reference](../tooluniverse-protein-design-campaign/references/ph-selectivity.md).209210**Output**: Ranked mutation hypotheses, combination strategy, computed metrics where supported, and explicit unverified affinity/conditional-binding claims. Do not invent KD ranges or fold improvements.211212See `WORKFLOW_DETAILS.md` Phase 4 for code examples.213214---215216## Phase 5: Developability Assessment217218**Goal**: Comprehensive developability scoring (0-100) across five dimensions.219220**Key steps**:2211. **Aggregation, pI, hydrophobic patches** (sequence-computable): run `scripts/developability.py --seq <VH/VL>`. It returns AGGRESCAN aggregation-prone regions (real per-residue propensity), isoelectric point, and Kyte-Doolittle hydrophobic patches.2222. **PTM liability**: Scan for deamidation (NG/NS), isomerization (DG/DS), oxidation (Met/Trp), N-glycosylation (N-X-S/T)2233. **Stability (Tm)** and **4. Expression titer**: NOT computable from sequence alone — use external predictors (sequence→Tm/titer ML tools; report only if actually run). Do not fabricate Tm/titer numbers.2245. **Solubility**: external solubility predictor (e.g. NetSolP)225226**Scoring**: Weighted average (aggregation 0.30, PTM 0.25, stability 0.20, expression 0.15, solubility 0.10).227Tiers: T1 (>75), T2 (60-75), T3 (<60).228229**Output**: Component scores, overall score, tier classification, mitigation recommendations.230231See `WORKFLOW_DETAILS.md` Phase 5 and `CHECKLISTS.md` for scoring details.232233---234235## Phase 6: Immunogenicity Prediction236237**Goal**: Predict immunogenicity risk and design deimmunization strategy.238239**Key steps**:2401. Scan 9-mer peptides against IEDB for MHC-II binding epitopes2412. Count non-human residues in framework regions2423. Assess aggregation-related immunogenicity2434. Calculate total risk score (0-100, lower is better): Low <30, Medium 30-60, High >602445. Propose deimmunization mutations (remove T-cell epitopes while preserving CDRs)2456. Compare with clinical precedent ADA rates246247**Output**: T-cell epitope list, risk score breakdown, deimmunization strategy, clinical comparison.248249See `WORKFLOW_DETAILS.md` Phase 6 for code examples.250251---252253## Phase 7: Manufacturing Feasibility254255**Goal**: Assess expression, purification, formulation, and CMC feasibility.256257**Key steps**:2581. Assess codon optimization for CHO, identify rare codons -- `CodonUsage_get_optimal_codons(taxid=<Cricetulus griseus NCBI taxid>)` gives the CHO-preferred codon per amino acid to compare against; resolve the taxid first if not already known rather than assuming one. `CodonUsage_get_table` (same taxid) additionally gives per-codon usage frequency, needed to flag which codons in the actual sequence are rare in CHO specifically -- human codon-usage tables (e.g. from tooluniverse-sequence-analysis) are the wrong reference here, rare-in-human is not the same set as rare-in-CHO2592. Design signal peptide2603. Plan 3-step purification: Protein A capture -> cation exchange polishing -> viral nanofiltration2614. Recommend formulation (buffer, pH, stabilizer, tonicity)2625. Define analytical characterization panel (SEC-MALS, CEX, CE-SDS, SPR, DSF)2636. Estimate CMC timeline and costs (typically 18-24 months, $1.65-2.65M to IND)264265**Output**: Expression assessment, purification strategy, formulation recommendation, CMC timeline.266267See `MANUFACTURING.md` for detailed manufacturing content and `WORKFLOW_DETAILS.md` Phase 7 for code.268269---270271## Phase 8: Final Report & Recommendations272273**Goal**: Compile all findings into a ranked recommendation with validation plan.274275**Key outputs**:2761. **Top candidate** with key metrics (humanness, affinity, developability, immunogenicity, stability, expression)2772. **Key improvements** table comparing original vs. optimized2783. **Experimental validation plan**: In vitro (3-4 months) -> Lead optimization (2-3 months) -> Preclinical (6-12 months)2794. **Backup variants** with profiles and recommendations2805. **IP considerations**: FTO analysis, CDR novelty, patentability2816. **Next steps**: Immediate (month 1-3), short-term (4-6), long-term (7-24)282283See `REPORT_TEMPLATE.md` for the full report template.284285---286287## Tool Reference288289### IMGT Tools290- `IMGT_search_genes`: Search germline genes (IGHV, IGKV, etc.)291- `IMGT_get_sequence`: Get germline sequences292- `IMGT_get_gene_info`: Database information293294### Antibody Databases295- `SAbDab_search_structures`: Search antibody structures296- `SAbDab_get_structure`: Get structure details297- `TheraSAbDab_search_therapeutics`: Search by name298- `TheraSAbDab_search_by_target`: Search by target antigen299300### Immunogenicity301- `iedb_search_epitopes`: Search epitopes302- `iedb_search_bcell`: B-cell epitopes303- `iedb_search_mhc`: MHC-II epitopes304- `iedb_get_epitope_references`: Citations305306### Structure & Target307- `ESMFold_predict_structure`: Fv structure prediction from a sequence308- `alphafold_get_prediction`: Retrieve a precomputed AlphaFold DB model by UniProt accession309- `UniProt_get_entry_by_accession`: Target info310- `RCSBData_get_entry`: Experimental structures311312### Reagent Verification (Antibody Registry)313- `AntibodyRegistry_search`: Find validated commercial antibodies for a target (by target gene/protein name, antibody name, or vendor keyword). Returns each hit's RRID, target (with UniProt/Entrez id), clonality, host organism, applications, vendor/catalog number, and defining citation. Useful during Phase 1 (target/precedent research) to see what antibody reagents already exist against a target antigen before designing a new one, or during Phase 8 (validation planning) to identify a benchmark reagent.314- `AntibodyRegistry_get_by_rrid`: Resolve a specific RRID (e.g. `AB_2298772`, accepts the bare numeric id or `RRID:AB_...` form too) cited in a methods section to its full record — verify which exact antibody a paper used rather than assuming from the name alone. Live-verified: both tools return real records (e.g. `AntibodyRegistry_search(query="GFAP")` -> a real GFAP rabbit mAb with UniProt target P14136).315- `SciCrunch_resolve_rrid`: General RRID resolver for everything that ISN'T an antibody — cell lines (`CVCL_` prefix), model organisms, software/tools (`SCR_` prefix), and plasmids. Use it when a clinical-precedent or validation paper's methods section also cites the cell line or software used alongside an antibody. It will still resolve an `AB_` RRID, but only returns generic fields (name, description, categories, proper_citation) — no target/clonality/host — so prefer `AntibodyRegistry_get_by_rrid` for antibody RRIDs specifically. Live-verified: `SciCrunch_resolve_rrid(rrid="CVCL_0045")` returns the real HEK293 cell-line record (12 synonyms, `proper_citation: "NCBI_Iran Cat# C497, RRID:CVCL_0045"`); on the same `AB_2298772` antibody RRID above it returns only `name`/`description`/`categories` with no target info, confirming the richer-antibody-fields distinction.316317### Systems Biology (for Bispecifics)318- `STRING_get_interaction_partners`: Protein interactions319- `STRING_get_enrichment`: Pathway analysis320321---322323## Reference Files324325| File | Contents |326|------|----------|327| `QUICK_START.md` | Getting started guide, SOAP tool parameters, Python SDK and MCP usage |328| `WORKFLOW_DETAILS.md` | Code examples for all 8 phases |329| `REPORT_TEMPLATE.md` | Full report template with section formats and example tables |330| `MANUFACTURING.md` | Detailed manufacturing content (expression, purification, formulation, CMC) |331| `EXAMPLES.md` | Complete clinical scenario examples (humanization, affinity, bispecific) |332| `CHECKLISTS.md` | Evidence grading, completeness checklists, scoring details, special considerations |