AlphaGenome Prompt Examples¶
This guide provides example prompts for using AlphaGenome through BioMCP to analyze genetic variants. These prompts are designed for use with AI assistants like Claude that have BioMCP integrated.
Basic Variant Analysis¶
Known Pathogenic Variants¶
BRAF V600E (Melanoma)
Use alphagenome_predictor to analyze the regulatory effects of the BRAF V600E mutation (chr7:140753336 A>T)
TP53 Hotspot Mutations
Predict how the TP53 R175H mutation (chr17:7675088 C>T) affects gene expression and chromatin accessibility
EGFR Resistance Mutation
Non-coding Variant Analysis¶
Promoter Variants
Enhancer Variants
UTR Variants
Research-Oriented Analysis¶
Variant Prioritization¶
Multiple Variant Screening
I have a list of variants from whole genome sequencing. Can you use alphagenome_predictor to analyze these and identify which ones likely have the strongest regulatory effects:
- chr3:178936091 G>A
- chr12:25398285 C>T
- chr19:11224301 G>T
Regulatory Impact Ranking
Analyze these non-coding variants and rank them by predicted regulatory impact:
1. chr5:1282543 C>T (TERT promoter)
2. chr8:128750412 A>G (MYC enhancer)
3. chr17:7571720 G>A (TP53 promoter)
Splicing Analysis¶
Intronic Variants
Use AlphaGenome to predict if this intronic variant affects splicing: chr2:215593426 A>G in the BARD1 gene
Splice Site Variants
Analyze these variants near splice sites for potential splicing alterations:
- chr11:108198135 C>T (ATM gene, +5 position)
- chr13:32340700 G>A (BRCA2 gene, -3 position)
Tissue-Specific Analysis¶
Breast Tissue Analysis
Liver-Specific Effects
Use alphagenome_predictor with liver tissue context (UBERON:0002107) for this variant: chr16:31356190 G>A in the FTO gene
Multi-Tissue Comparison
Compare the effects of chr12:25398285 C>T across:
- Brain tissue (UBERON:0000955)
- Liver tissue (UBERON:0002107)
- Lung tissue (UBERON:0002048)
Clinical Research Workflows¶
Variant of Uncertain Significance (VUS) Analysis¶
Complete VUS Workup
I found a variant of uncertain significance in a cancer patient: chr9:21971076 C>T in CDKN2A.
1. First use variant_getter to see known annotations
2. Then use alphagenome_predictor to assess potential regulatory impacts
3. Search for articles about similar CDKN2A variants
Pharmacogenomics¶
Drug Metabolism Variants
Analyze how the CYP2D6 variant chr22:42130692 G>A might affect drug metabolism gene expression using AlphaGenome
Warfarin Sensitivity
Rare Disease Investigation¶
Mitochondrial Disease
This patient has a rare variant chr15:89859516 C>T in the POLG gene. Use alphagenome_predictor to understand if it might affect mitochondrial DNA polymerase expression
Neurodevelopmental Disorders
Analyze this de novo variant in a child with developmental delay: chr2:166199235 C>G in the SCN1A gene
Comparative Analysis¶
Multiple Variant Comparison¶
Gene-Wide Analysis
Compare the predicted regulatory effects of these three BRCA1 variants using AlphaGenome:
- chr17:41245237 G>A (promoter)
- chr17:41244936 G>A (5' UTR)
- chr17:41243451 T>C (intron 2)
Hotspot Comparison
I'm studying why some TP53 mutations are more severe than others. Use alphagenome_predictor to compare these hotspot mutations:
- R175H (chr17:7675088 C>T)
- R248W (chr17:7674220 G>A)
- R273H (chr17:7673802 C>T)
Allele-Specific Analysis¶
Alternative Alleles
This GWAS hit is at chr5:1280000. Use alphagenome_predictor to analyze all possible variants:
- chr5:1280000 A>G
- chr5:1280000 A>C
- chr5:1280000 A>T
Which alternate allele has the strongest predicted effect?
Advanced Research Prompts¶
Long-Range Regulatory Analysis¶
Extended Window Analysis
Use alphagenome_predictor with --interval 1048576 to analyze long-range regulatory effects of chr8:128750000 A>G near the MYC oncogene
TAD Boundary Variants
Compound Heterozygote Analysis¶
Trans Configuration
Analyze these two variants in trans in the same gene:
- Maternal: chr11:47342697 C>T (MYBPC3)
- Paternal: chr11:47380142 G>A (MYBPC3)
What are their individual regulatory effects?
Cancer Research¶
Driver vs Passenger
Help distinguish driver from passenger mutations. Analyze these variants from a tumor:
1. chr7:140753336 A>T (BRAF V600E)
2. chr3:41266101 C>T (CTNNB1 S33F)
3. chr1:115256529 G>A (NRAS Q61R)
Which show the strongest regulatory effects?
Tumor Suppressor Analysis
Analyze non-coding variants near tumor suppressors:
- chr17:7565097 C>T (TP53 promoter)
- chr13:32316461 A>G (BRCA2 promoter)
- chr17:41196312 G>C (BRCA1 promoter)
Integration with Other BioMCP Tools¶
Full Variant Characterization¶
Literature + Prediction
1. Search for articles about KRAS G12D mutations using article_searcher
2. Then use alphagenome_predictor to analyze chr12:25245350 C>T
3. Compare the literature findings with AlphaGenome predictions
Database + AI Analysis
1. Use variant_searcher to find pathogenic variants in the BRCA2 gene
2. Pick the top 3 results
3. Analyze each with alphagenome_predictor
4. Which has the strongest predicted regulatory impact?
Clinical Trial Context¶
Treatment Target Analysis
1. Find clinical trials for BRAF V600E melanoma
2. Use AlphaGenome to understand why this mutation (chr7:140753336 A>T) is so impactful
3. Does the regulatory effect explain the treatment response?
Tips for Effective Prompts¶
Required Information¶
- Chromosome: Use "chr" prefix (e.g., chr7, chrX)
- Position: 1-based coordinate from reference genome
- Reference allele: Current base(s) at that position
- Alternate allele: Changed base(s)
Optional Parameters¶
- Interval size: 2048, 16384, 131072, 524288, or 1048576
- Tissue type: UBERON ontology terms
- Multiple variants: Analyze in single prompt for comparison
Best Practices¶
- Be specific - Include exact coordinates
- Provide context - Mention gene names and known effects
- Ask for interpretation - Request specific insights
- Combine tools - Use multiple BioMCP tools for comprehensive analysis
- Consider mechanism - Ask about expression, splicing, or chromatin
Example Multi-Step Workflow¶
I'm investigating a patient with suspected hereditary cancer syndrome. They have these variants:
1. First, check each variant in databases:
- Use variant_getter on chr17:41245237 G>A
- Use variant_getter on chr13:32340700 G>A
- Use variant_getter on chr11:108198135 C>T
2. Then predict regulatory effects:
- Use alphagenome_predictor on each variant
- Compare which has the strongest impact
3. Search literature:
- Find articles about each affected gene
- Look for similar cases
4. Summarize findings:
- Which variant is most likely pathogenic?
- What functional evidence supports this?
This structured approach combines AlphaGenome's predictive power with BioMCP's database access for comprehensive variant analysis.