File contents name: 'tcr-pmhc-prediction-agent'
description: 'AI-powered TCR-peptide-MHC interaction prediction using AlphaFold3 and deep learning for therapeutic TCR discovery, neoantigen validation, and T cell immunogenicity assessment.'
measurable_outcome: Execute skill workflow successfully with valid output within 15 minutes.
allowed-tools:
- read_file
- run_shell_command
TCR-pMHC Prediction Agent
The TCR-pMHC Prediction Agent predicts T-cell receptor interactions with peptide-MHC complexes using AlphaFold3-based structural modeling and deep learning. Accurate TCR-pMHC prediction enables therapeutic TCR discovery, neoantigen vaccine validation, and identification of immunogenic epitopes for cancer and infectious disease applications.
When to Use This Skill
When predicting which peptides a TCR will recognize.
For validating neoantigen immunogenicity computationally.
To screen therapeutic TCR candidates against target antigens.
When assessing cross-reactivity of TCRs with self-peptides.
For understanding TCR specificity determinants.
Core Capabilities
Binding Prediction : Predict TCR-pMHC binding affinity/probability.
Structural Modeling : Generate TCR-pMHC complex structures with AlphaFold3.
Epitope Specificity : Determine which epitopes a TCR recognizes.
Cross-Reactivity Assessment : Predict off-target self-peptide binding.
Immunogenicity Scoring : Rank peptide immunogenicity.
Therapeutic TCR Screening : Screen TCRs for desired specificity.
Prediction Approaches
Approach
Method
Strengths
AlphaFold3
Structure prediction
High accuracy, interpretable
TCR-BERT
Sequence transformer
Fast, large-scale
ERGO-II
RNN-based
Established benchmark
pMTnet
Multi-task learning
Generalizable
NetTCR
CNN-based
HLA-specific
TITAN
Attention-based
State-of-art sequence
Workflow
Input : TCR sequence (alpha/beta CDR3), peptide, HLA allele.
Structure Prediction : Generate pMHC and TCR structures.
Docking : Model TCR-pMHC complex.
Scoring : Calculate binding probability/affinity.
Cross-Reactivity : Screen against self-peptide database.
Validation Features : Extract structural determinants.
Output : Binding predictions, structures, safety assessment.
Example Usage
User : "Predict whether this tumor-reactive TCR binds the identified neoantigen and check for cross-reactivity with self-peptides."
Agent Action :
python3 Skills/Immunology_Vaccines/TCR_pMHC_Prediction_Agent/tcr_pmhc_predict.py \
--tcr_alpha_cdr3 CAVSDRGSTLGRLYF \
--tcr_beta_cdr3 CASSLGQAYEQYF \
--tcr_v_genes TRAV12-1,TRBV7-9 \
--peptide KRAS_G12D_VVGADGVGK \
--hla HLA-A*11:01 \
--check_cross_reactivity true \
--self_peptide_db human_proteome_9mers.fasta \
--method alphafold3 \
--output tcr_pmhc_results/
Input Requirements
Input
Format
Required
TCR CDR3 alpha
Amino acid sequence
Yes
TCR CDR3 beta
Amino acid sequence
Yes
V gene usage
IMGT notation
Recommended
Peptide
8-11mer amino acids
Yes
HLA allele
4-digit resolution
Yes
Output Components
Output
Description
Format
Binding Score
Probability of binding
.json
Complex Structure
TCR-pMHC model
.pdb
Contact Map
Residue interactions
.csv, .png
Cross-Reactivity
Self-peptide hits
.csv
Confidence Score
Prediction reliability
.json
Binding Determinants
Key residues
.csv
AlphaFold3 Integration
Component
Application
Output
pMHC Modeling
Peptide-MHC structure
Complex structure
TCR Modeling
Variable region structure
TCR structure
Complex Prediction
Full ternary complex
Docked model
pLDDT Scores
Confidence per residue
Quality metric
PAE
Positional error
Interface confidence
Binding Prediction Thresholds
Score Range
Interpretation
Action
>0.9
Strong predicted binder
High confidence
0.7-0.9
Moderate predicted binder
Likely positive
0.5-0.7
Weak/uncertain
Experimental validation needed
<0.5
Predicted non-binder
Low priority
AI/ML Components
Structural Prediction :
AlphaFold3 for complex modeling
Molecular dynamics refinement
Interface scoring functions
Sequence Models :
TCR-specific language models
Cross-attention for TCR-peptide
Transfer learning from pMHC binding
Cross-Reactivity :
Embedding similarity search
Structural hotspot analysis
Self-tolerance modeling
Performance Benchmarks
Method
Dataset
AUC
Notes
AlphaFold3
VDJdb benchmark
0.85
Structural
TCR-BERT
IEDB
0.82
Fast screening
ERGO-II
McPAS-TCR
0.78
Established
Ensemble
Combined
0.88
Best overall
Clinical Applications
Application
Use Case
TCR-pMHC Role
Neoantigen Vaccines
Validate immunogenicity
Predict T cell response
TCR-T Therapy
Select therapeutic TCRs
Screen candidates
Safety Assessment
Check cross-reactivity
Avoid autoimmunity
Epitope Discovery
Find immunogenic peptides
Prioritize targets
Prerequisites
Python 3.10+
AlphaFold3 installation
PyTorch, transformers
BioPython, MDAnalysis
GPU with 16GB+ VRAM
Self-peptide reference database
Related Skills
TCR_Repertoire_Analysis_Agent - Repertoire analysis
Neoantigen_Prediction_Agent - Neoantigen identification
HLA_Typing_Agent - HLA determination
CART_Design_Optimizer_Agent - TCR-based therapy
Cross-Reactivity Safety Analysis
Database
Content
Purpose
Human Proteome
All self-peptides
Primary safety
Tissue-Specific
Expression-weighted
Toxicity prediction
Viral Mimicry
Viral homologs
Infection mimics
Cancer-Testis
CT antigens
On-target activity
Structural Determinants
Feature
Location
Significance
CDR3 beta apex
Peptide contact
Specificity
CDR3 alpha
MHC/peptide
Fine-tuning
CDR1/2
MHC helices
HLA restriction
Germline-encoded
Framework
Base recognition
Special Considerations
HLA Restriction : Predictions are HLA-specificCDR3 Dominance : CDR3 beta often most predictivePaired Chains : Alpha-beta pairing crucialStructural Validation : Validate with known structuresExperimental Follow-up : Tetramer/functional validation
Limitations
Limitation
Impact
Mitigation
Training Data Bias
Common HLA over-represented
Use diverse training
Novel TCRs
Out-of-distribution
Lower confidence
Post-translational
PTM peptides not modeled
Experimental validation
Dynamics
Static structures
MD simulation
Author
AI Group - Biomedical AI Platform
1 --- 2 name: tcr-pmhc-prediction-agent 3 description: <!-- 4 --- 5 <!-- 6 # COPYRIGHT NOTICE 7 # This file is part of the "Universal Biomedical Skills" project. 8 # Copyright (c) 2026 MD BABU MIA, PhD <md.babu.mia@mssm.edu> 9 # All Rights Reserved. 10 # 11 # This code is proprietary and confidential. 12 # Unauthorized copying of this file, via any medium is strictly prohibited. 13 # 14 # Provenance: Authenticated by MD BABU MIA 15 16 --> 17 18 --- 19 name: 'tcr-pmhc-prediction-agent' 20 description: 'AI-powered TCR-peptide-MHC interaction prediction using AlphaFold3 and deep learning for therapeutic TCR discovery, neoantigen validation, and T cell immunogenicity assessment.' 21 measurable_outcome: Execute skill workflow successfully with valid output within 15 minutes. 22 allowed-tools: 23 - read_file 24 - run_shell_command 25 --- 26 27 28 # TCR-pMHC Prediction Agent 29 30 The **TCR-pMHC Prediction Agent** predicts T-cell receptor interactions with peptide-MHC complexes using AlphaFold3-based structural modeling and deep learning. Accurate TCR-pMHC prediction enables therapeutic TCR discovery, neoantigen vaccine validation, and identification of immunogenic epitopes for cancer and infectious disease applications. 31 32 ## When to Use This Skill 33 34 * When predicting which peptides a TCR will recognize. 35 * For validating neoantigen immunogenicity computationally. 36 * To screen therapeutic TCR candidates against target antigens. 37 * When assessing cross-reactivity of TCRs with self-peptides. 38 * For understanding TCR specificity determinants. 39 40 ## Core Capabilities 41 42 1. **Binding Prediction**: Predict TCR-pMHC binding affinity/probability. 43 44 2. **Structural Modeling**: Generate TCR-pMHC complex structures with AlphaFold3. 45 46 3. **Epitope Specificity**: Determine which epitopes a TCR recognizes. 47 48 4. **Cross-Reactivity Assessment**: Predict off-target self-peptide binding. 49 50 5. **Immunogenicity Scoring**: Rank peptide immunogenicity. 51 52 6. **Therapeutic TCR Screening**: Screen TCRs for desired specificity. 53 54 ## Prediction Approaches 55 56 | Approach | Method | Strengths | 57 |----------|--------|-----------| 58 | AlphaFold3 | Structure prediction | High accuracy, interpretable | 59 | TCR-BERT | Sequence transformer | Fast, large-scale | 60 | ERGO-II | RNN-based | Established benchmark | 61 | pMTnet | Multi-task learning | Generalizable | 62 | NetTCR | CNN-based | HLA-specific | 63 | TITAN | Attention-based | State-of-art sequence | 64 65 ## Workflow 66 67 1. **Input**: TCR sequence (alpha/beta CDR3), peptide, HLA allele. 68 69 2. **Structure Prediction**: Generate pMHC and TCR structures. 70 71 3. **Docking**: Model TCR-pMHC complex. 72 73 4. **Scoring**: Calculate binding probability/affinity. 74 75 5. **Cross-Reactivity**: Screen against self-peptide database. 76 77 6. **Validation Features**: Extract structural determinants. 78 79 7. **Output**: Binding predictions, structures, safety assessment. 80 81 ## Example Usage 82 83 **User**: "Predict whether this tumor-reactive TCR binds the identified neoantigen and check for cross-reactivity with self-peptides." 84 85 **Agent Action**: 86 ```bash 87 python3 Skills/Immunology_Vaccines/TCR_pMHC_Prediction_Agent/tcr_pmhc_predict.py \ 88 --tcr_alpha_cdr3 CAVSDRGSTLGRLYF \ 89 --tcr_beta_cdr3 CASSLGQAYEQYF \ 90 --tcr_v_genes TRAV12-1,TRBV7-9 \ 91 --peptide KRAS_G12D_VVGADGVGK \ 92 --hla HLA-A*11:01 \ 93 --check_cross_reactivity true \ 94 --self_peptide_db human_proteome_9mers.fasta \ 95 --method alphafold3 \ 96 --output tcr_pmhc_results/ 97 ``` 98 99 ## Input Requirements 100 101 | Input | Format | Required | 102 |-------|--------|----------| 103 | TCR CDR3 alpha | Amino acid sequence | Yes | 104 | TCR CDR3 beta | Amino acid sequence | Yes | 105 | V gene usage | IMGT notation | Recommended | 106 | Peptide | 8-11mer amino acids | Yes | 107 | HLA allele | 4-digit resolution | Yes | 108 109 ## Output Components 110 111 | Output | Description | Format | 112 |--------|-------------|--------| 113 | Binding Score | Probability of binding | .json | 114 | Complex Structure | TCR-pMHC model | .pdb | 115 | Contact Map | Residue interactions | .csv, .png | 116 | Cross-Reactivity | Self-peptide hits | .csv | 117 | Confidence Score | Prediction reliability | .json | 118 | Binding Determinants | Key residues | .csv | 119 120 ## AlphaFold3 Integration 121 122 | Component | Application | Output | 123 |-----------|-------------|--------| 124 | pMHC Modeling | Peptide-MHC structure | Complex structure | 125 | TCR Modeling | Variable region structure | TCR structure | 126 | Complex Prediction | Full ternary complex | Docked model | 127 | pLDDT Scores | Confidence per residue | Quality metric | 128 | PAE | Positional error | Interface confidence | 129 130 ## Binding Prediction Thresholds 131 132 | Score Range | Interpretation | Action | 133 |-------------|----------------|--------| 134 | >0.9 | Strong predicted binder | High confidence | 135 | 0.7-0.9 | Moderate predicted binder | Likely positive | 136 | 0.5-0.7 | Weak/uncertain | Experimental validation needed | 137 | <0.5 | Predicted non-binder | Low priority | 138 139 ## AI/ML Components 140 141 **Structural Prediction**: 142 - AlphaFold3 for complex modeling 143 - Molecular dynamics refinement 144 - Interface scoring functions 145 146 **Sequence Models**: 147 - TCR-specific language models 148 - Cross-attention for TCR-peptide 149 - Transfer learning from pMHC binding 150 151 **Cross-Reactivity**: 152 - Embedding similarity search 153 - Structural hotspot analysis 154 - Self-tolerance modeling 155 156 ## Performance Benchmarks 157 158 | Method | Dataset | AUC | Notes | 159 |--------|---------|-----|-------| 160 | AlphaFold3 | VDJdb benchmark | 0.85 | Structural | 161 | TCR-BERT | IEDB | 0.82 | Fast screening | 162 | ERGO-II | McPAS-TCR | 0.78 | Established | 163 | Ensemble | Combined | 0.88 | Best overall | 164 165 ## Clinical Applications 166 167 | Application | Use Case | TCR-pMHC Role | 168 |-------------|----------|---------------| 169 | Neoantigen Vaccines | Validate immunogenicity | Predict T cell response | 170 | TCR-T Therapy | Select therapeutic TCRs | Screen candidates | 171 | Safety Assessment | Check cross-reactivity | Avoid autoimmunity | 172 | Epitope Discovery | Find immunogenic peptides | Prioritize targets | 173 174 ## Prerequisites 175 176 * Python 3.10+ 177 * AlphaFold3 installation 178 * PyTorch, transformers 179 * BioPython, MDAnalysis 180 * GPU with 16GB+ VRAM 181 * Self-peptide reference database 182 183 ## Related Skills 184 185 * TCR_Repertoire_Analysis_Agent - Repertoire analysis 186 * Neoantigen_Prediction_Agent - Neoantigen identification 187 * HLA_Typing_Agent - HLA determination 188 * CART_Design_Optimizer_Agent - TCR-based therapy 189 190 ## Cross-Reactivity Safety Analysis 191 192 | Database | Content | Purpose | 193 |----------|---------|---------| 194 | Human Proteome | All self-peptides | Primary safety | 195 | Tissue-Specific | Expression-weighted | Toxicity prediction | 196 | Viral Mimicry | Viral homologs | Infection mimics | 197 | Cancer-Testis | CT antigens | On-target activity | 198 199 ## Structural Determinants 200 201 | Feature | Location | Significance | 202 |---------|----------|--------------| 203 | CDR3 beta apex | Peptide contact | Specificity | 204 | CDR3 alpha | MHC/peptide | Fine-tuning | 205 | CDR1/2 | MHC helices | HLA restriction | 206 | Germline-encoded | Framework | Base recognition | 207 208 ## Special Considerations 209 210 1. **HLA Restriction**: Predictions are HLA-specific 211 2. **CDR3 Dominance**: CDR3 beta often most predictive 212 3. **Paired Chains**: Alpha-beta pairing crucial 213 4. **Structural Validation**: Validate with known structures 214 5. **Experimental Follow-up**: Tetramer/functional validation 215 216 ## Limitations 217 218 | Limitation | Impact | Mitigation | 219 |------------|--------|------------| 220 | Training Data Bias | Common HLA over-represented | Use diverse training | 221 | Novel TCRs | Out-of-distribution | Lower confidence | 222 | Post-translational | PTM peptides not modeled | Experimental validation | 223 | Dynamics | Static structures | MD simulation | 224 225 ## Author 226 227 AI Group - Biomedical AI Platform 228 229 230 <!-- AUTHOR_SIGNATURE: 9a7f3c2e-MD-BABU-MIA-2026-MSSM-SECURE -->
commit 1c93a8f5a7
Frequently asked questions How do I install the Tcr Pmhc Prediction Agent skill? Run npx skillmds@latest add biotender-max/tcr-pmhc-prediction-agent in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.
What does the Tcr Pmhc Prediction Agent skill do? <!-- It is listed under AI & ML on SkillMD.
Is Tcr Pmhc Prediction Agent safe to use? This skill has not completed SkillMD's automated safety review yet. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.
Which AI agents work with Tcr Pmhc Prediction Agent? This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.
Is Tcr Pmhc Prediction Agent free to use? Yes. Installing skills from SkillMD is free, and the skill stays under its author's original license.
Who published Tcr Pmhc Prediction Agent? BioTender-max (@biotender-max) published this skill. Their other Agent Skills are listed on their SkillMD profile.
BioTender-max