---name: pharmacogenomics-agent
description: AI-powered pharmacogenomic analysis for drug response prediction, adverse event risk assessment, and precision dosing using multi-omics data and deep learning models.
license: MIT
metadata:
author: AI Group
version: "1.0.0"
created: "2026-01-19"
compatibility:
- system: Python 3.10+
allowed-tools:
- run_shell_command
- read_file
- write_file
keywords:
- pharmacogenomics-agent
- automation
- biomedical
measurable_outcome: execute task with >95% success rate.
---"
Pharmacogenomics Agent
The Pharmacogenomics Agent integrates AI and multi-omics data to predict individual drug responses, optimize medication dosing, and minimize adverse events. It implements CPIC guidelines while leveraging deep learning for complex polygenic drug response phenotypes.
When to Use This Skill
- When interpreting pharmacogenomic variants (CYP450, HLA, transporters) for drug selection.
- To predict drug response using transcriptomic and proteomic biomarkers.
- For calculating polygenic risk scores for drug efficacy/toxicity.
- When optimizing doses for narrow therapeutic index drugs.
- To identify drug-drug-gene interactions.
Core Capabilities
Variant Interpretation: Translates star allele genotypes (*1/*2) into metabolizer phenotypes and actionable CPIC recommendations.
Multi-Omics Response Prediction: Deep learning models (DeepDRA, MOViDA) integrate genomic, transcriptomic, and proteomic features for drug response prediction.
Polygenic Risk Scoring: Combines effects of thousands of variants to stratify patients beyond single-gene pharmacogenomics.
Adverse Event Prediction: Identifies genetic risk factors for serious adverse reactions (HLA associations, G6PD deficiency).
Dose Optimization: AI-guided dosing for warfarin, tacrolimus, fluoropyrimidines, thiopurines, and other PGx-guided drugs.
Drug-Drug-Gene Interactions: Detects complex interactions where genetic variants modify drug interaction severity.
CPIC-Guided Genes and Drugs
| Gene |
Drugs |
Clinical Impact |
| CYP2D6 |
Codeine, tamoxifen, antidepressants |
Metabolizer status affects efficacy/toxicity |
| CYP2C19 |
Clopidogrel, PPIs, antidepressants |
Loss-of-function affects activation |
| CYP2C9/VKORC1 |
Warfarin |
Dose requirements vary 10-fold |
| TPMT/NUDT15 |
Thiopurines |
Myelosuppression risk |
| DPYD |
Fluoropyrimidines |
Severe/fatal toxicity in deficient patients |
| HLA-B*57:01 |
Abacavir |
Hypersensitivity screening |
| HLA-B*15:02 |
Carbamazepine |
SJS/TEN risk in Asian populations |
Workflow
Input: Patient genotype data (VCF, genotyping array), medication list, clinical parameters.
Star Allele Calling: Translate variants to star alleles using Stargazer or PharmCAT.
Phenotype Assignment: Determine metabolizer status (PM, IM, NM, UM) for each gene.
Guideline Lookup: Retrieve CPIC/DPWG recommendations for patient's medications.
Multi-Omics Prediction: Apply deep learning for complex response phenotypes.
Output: Drug-specific recommendations, dose adjustments, alternative medications, interaction alerts.
Example Usage
User: "Interpret this patient's pharmacogenomic panel and provide recommendations for their current medications."
Agent Action:
python3 Skills/Precision_Medicine/Pharmacogenomics_Agent/pgx_analyzer.py \
--genotype patient_pgx_panel.vcf \
--medications current_meds.json \
--guidelines cpic_dpwg \
--risk_scores oncology_response \
--output pgx_recommendations.json
AI Models for Drug Response
| Model |
Architecture |
Application |
Performance |
| DeepDRA |
Autoencoders |
Drug response from transcriptomics |
AUC 0.99 |
| MOViDA |
Multi-omics VAE |
Interpretable response prediction |
State-of-art |
| DrugCell |
Graph neural network |
Drug synergy prediction |
Improved over baselines |
| PaccMann |
Multimodal attention |
Cancer drug sensitivity |
Clinical translation |
Polygenic Drug Response
Beyond single-gene PGx, polygenic scores capture:
- Efficacy polygenic scores: Statin LDL response, antidepressant remission
- Toxicity polygenic scores: Metformin GI intolerance, opioid dependence risk
- Combined scores: Integrating PRS with PGx for personalized prediction
Prerequisites
- Python 3.10+
- PharmCAT or Stargazer for star allele calling
- CPIC/DPWG guideline databases
- Deep learning frameworks (PyTorch)
- Optional: Expression data for multi-omics models
Related Skills
- Variant_Interpretation - For general variant classification
- Drug_Repurposing - For alternative drug identification
- Clinical_Trials - For PGx-guided trial matching
Implementation Notes
Clinical Integration:
- Returns structured FHIR-compatible recommendations
- Supports CDS Hooks for real-time EMR alerts
- Audit trail for clinical decision support
Quality Metrics:
- Validated against PharmGKB annotations
- Concordance with reference laboratory calls
- Regular updates with new CPIC guidelines
Author
AI Group - Biomedical AI Platform
1---2name: pharmacogenomics-agent3description: ---name: pharmacogenomics-agent4---5
6---name: pharmacogenomics-agent
7description: AI-powered pharmacogenomic analysis for drug response prediction, adverse event risk assessment, and precision dosing using multi-omics data and deep learning models.
8license: MIT
9metadata:
10 author: AI Group
11 version: "1.0.0"
12 created: "2026-01-19"
13compatibility:
14 - system: Python 3.10+
15allowed-tools:
16 - run_shell_command
17 - read_file
18 - write_file
19
20keywords:
21 - pharmacogenomics-agent
22 - automation
23 - biomedical
24measurable_outcome: execute task with >95% success rate.
25---"
26
27# Pharmacogenomics Agent
28
29The **Pharmacogenomics Agent** integrates AI and multi-omics data to predict individual drug responses, optimize medication dosing, and minimize adverse events. It implements CPIC guidelines while leveraging deep learning for complex polygenic drug response phenotypes.
30
31## When to Use This Skill
32
33* When interpreting pharmacogenomic variants (CYP450, HLA, transporters) for drug selection.
34* To predict drug response using transcriptomic and proteomic biomarkers.
35* For calculating polygenic risk scores for drug efficacy/toxicity.
36* When optimizing doses for narrow therapeutic index drugs.
37* To identify drug-drug-gene interactions.
38
39## Core Capabilities
40
411. **Variant Interpretation**: Translates star allele genotypes (*1/*2) into metabolizer phenotypes and actionable CPIC recommendations.
42
432. **Multi-Omics Response Prediction**: Deep learning models (DeepDRA, MOViDA) integrate genomic, transcriptomic, and proteomic features for drug response prediction.
44
453. **Polygenic Risk Scoring**: Combines effects of thousands of variants to stratify patients beyond single-gene pharmacogenomics.
46
474. **Adverse Event Prediction**: Identifies genetic risk factors for serious adverse reactions (HLA associations, G6PD deficiency).
48
495. **Dose Optimization**: AI-guided dosing for warfarin, tacrolimus, fluoropyrimidines, thiopurines, and other PGx-guided drugs.
50
516. **Drug-Drug-Gene Interactions**: Detects complex interactions where genetic variants modify drug interaction severity.
52
53## CPIC-Guided Genes and Drugs
54
55| Gene | Drugs | Clinical Impact |
56|------|-------|-----------------|
57| CYP2D6 | Codeine, tamoxifen, antidepressants | Metabolizer status affects efficacy/toxicity |
58| CYP2C19 | Clopidogrel, PPIs, antidepressants | Loss-of-function affects activation |
59| CYP2C9/VKORC1 | Warfarin | Dose requirements vary 10-fold |
60| TPMT/NUDT15 | Thiopurines | Myelosuppression risk |
61| DPYD | Fluoropyrimidines | Severe/fatal toxicity in deficient patients |
62| HLA-B*57:01 | Abacavir | Hypersensitivity screening |
63| HLA-B*15:02 | Carbamazepine | SJS/TEN risk in Asian populations |
64
65## Workflow
66
671. **Input**: Patient genotype data (VCF, genotyping array), medication list, clinical parameters.
68
692. **Star Allele Calling**: Translate variants to star alleles using Stargazer or PharmCAT.
70
713. **Phenotype Assignment**: Determine metabolizer status (PM, IM, NM, UM) for each gene.
72
734. **Guideline Lookup**: Retrieve CPIC/DPWG recommendations for patient's medications.
74
755. **Multi-Omics Prediction**: Apply deep learning for complex response phenotypes.
76
776. **Output**: Drug-specific recommendations, dose adjustments, alternative medications, interaction alerts.
78
79## Example Usage
80
81**User**: "Interpret this patient's pharmacogenomic panel and provide recommendations for their current medications."
82
83**Agent Action**:
84```bash
85python3 Skills/Precision_Medicine/Pharmacogenomics_Agent/pgx_analyzer.py \
86 --genotype patient_pgx_panel.vcf \
87 --medications current_meds.json \
88 --guidelines cpic_dpwg \
89 --risk_scores oncology_response \
90 --output pgx_recommendations.json
91```
92
93## AI Models for Drug Response
94
95| Model | Architecture | Application | Performance |
96|-------|--------------|-------------|-------------|
97| DeepDRA | Autoencoders | Drug response from transcriptomics | AUC 0.99 |
98| MOViDA | Multi-omics VAE | Interpretable response prediction | State-of-art |
99| DrugCell | Graph neural network | Drug synergy prediction | Improved over baselines |
100| PaccMann | Multimodal attention | Cancer drug sensitivity | Clinical translation |
101
102## Polygenic Drug Response
103
104Beyond single-gene PGx, polygenic scores capture:
105- **Efficacy polygenic scores**: Statin LDL response, antidepressant remission
106- **Toxicity polygenic scores**: Metformin GI intolerance, opioid dependence risk
107- **Combined scores**: Integrating PRS with PGx for personalized prediction
108
109## Prerequisites
110
111* Python 3.10+
112* PharmCAT or Stargazer for star allele calling
113* CPIC/DPWG guideline databases
114* Deep learning frameworks (PyTorch)
115* Optional: Expression data for multi-omics models
116
117## Related Skills
118
119* Variant_Interpretation - For general variant classification
120* Drug_Repurposing - For alternative drug identification
121* Clinical_Trials - For PGx-guided trial matching
122
123## Implementation Notes
124
125**Clinical Integration**:
126- Returns structured FHIR-compatible recommendations
127- Supports CDS Hooks for real-time EMR alerts
128- Audit trail for clinical decision support
129
130**Quality Metrics**:
131- Validated against PharmGKB annotations
132- Concordance with reference laboratory calls
133- Regular updates with new CPIC guidelines
134
135## Author
136
137AI Group - Biomedical AI Platform
majiayu000