chromosomal-instability-agent

---

name: 'chromosomal-instability-agent' description: 'AI-powered analysis of chromosomal instability (CIN) signatures for cancer prognosis, immunotherapy response prediction, and therapeutic vulnerability identification.' measurable_outcome: Execute skill workflow successfully with valid output within 15 minutes. allowed-tools:

• read_file
• run_shell_command

---

Chromosomal Instability Agent

The Chromosomal Instability Agent analyzes CIN signatures to predict cancer prognosis, immunotherapy response, and therapeutic vulnerabilities. It integrates copy number alterations, aneuploidy scores, and CIN-related gene expression for comprehensive genomic instability assessment.

When to Use This Skill

• When assessing tumor aneuploidy and chromosomal instability levels.
• To predict prognosis based on CIN signatures.
• For identifying tumors vulnerable to CIN-targeted therapies (PARP, ATR, WEE1).
• When analyzing immune evasion mechanisms related to CIN.
• To stratify patients for immunotherapy based on CIN status.

Core Capabilities

1. CIN Scoring: Calculate comprehensive CIN scores from copy number data.

2. Aneuploidy Quantification: Measure arm-level and focal copy number alterations.

3. CIN Gene Expression: Analyze CIN70 and other transcriptional signatures.

4. Immune Correlation: Assess CIN-immune microenvironment relationships.

5. Therapeutic Vulnerability: Identify CIN-targeted treatment options.

6. Prognostic Modeling: Predict outcomes based on CIN signatures.

CIN Metrics

Metric	Calculation	Interpretation
Aneuploidy score	Arm-level alterations	Chromosome-level CIN
SCNA burden	Total CNV alterations	Overall instability
Weighted GII	Fraction altered genome	Focal vs broad changes
CIN70	70-gene signature	Transcriptional CIN
WGII	Weighted genome instability	Comprehensive score

CIN70 Signature Genes

Core genes reflecting CIN phenotype:

• Mitotic checkpoint: BUB1, BUBR1, MAD2L1
• Kinetochore: CENPA, CENPF, NDC80
• DNA replication: MCM2-7, ORC1
• Cell cycle: CCNB1, CCNB2, CDK1, PLK1
• Chromosome segregation: AURKB, KIF2C, KIF11

Workflow

1. Input: Copy number data (segments), gene expression, mutation data.

2. CNV Analysis: Calculate arm-level and focal alterations.

3. Signature Scoring: Compute CIN70 and other transcriptional signatures.

4. Integration: Combine DNA and RNA-based CIN metrics.

5. Immune Analysis: Correlate CIN with TME composition.

6. Vulnerability Assessment: Identify targetable dependencies.

7. Output: CIN scores, prognosis, treatment recommendations.

Example Usage

User: "Analyze chromosomal instability in this breast cancer sample and identify treatment vulnerabilities."

Agent Action:

python3 Skills/Oncology/Chromosomal_Instability_Agent/cin_analyzer.py \
    --cnv_segments tumor_cnv.tsv \
    --expression rnaseq_tpm.tsv \
    --mutations somatic.maf \
    --tumor_type breast_cancer \
    --signatures cin70,cin25 \
    --output cin_report/

CIN and Immune Evasion

High CIN Associates With:

• Reduced immune infiltration
• Lower checkpoint inhibitor response
• Increased immune evasion
• cGAS-STING activation (paradoxical)

Mechanisms:

1. Loss of tumor suppressors on chromosome arms
2. Chronic inflammatory signaling
3. Aneuploidy-induced stress responses
4. Subclonal diversification

Therapeutic Vulnerabilities

Target	Agents	CIN Context
PARP	Olaparib, etc.	High CIN + HRD
ATR	Berzosertib	Replication stress
WEE1	Adavosertib	G2/M dependency
CHK1	Prexasertib	Cell cycle checkpoint
KIF11	Ispinesib	Mitotic dependency
Aurora kinases	Alisertib	Mitotic errors

CIN-Based Patient Stratification

CIN Level	Prognosis	ICI Response	Alternative Therapy
Low	Better	Better	Standard care
Intermediate	Variable	Variable	Combination therapy
High	Poor	Poor	CIN-targeted agents
Extreme	Very poor	Immune desert	Chemotherapy

AI/ML Components

CIN Score Prediction:

• Random forest on CNV features
• Expression-based CIN inference
• Multi-modal integration

Prognosis Modeling:

• Cox regression with CIN features
• Cancer-type specific models
• Integration with clinical variables

Therapeutic Matching:

• GDSC/CCLE drug sensitivity
• CIN-drug response correlations
• Combination predictions

Pan-Cancer CIN Patterns

Cancer Type	Typical CIN Level	Driver Events
Ovarian HGSOC	Very high	TP53, BRCA
Triple-neg breast	High	TP53, PI3K
Colorectal MSS	Moderate-high	APC, TP53
Colorectal MSI	Low	MMR deficiency
Thyroid (PTC)	Low	BRAF, RAS
Melanoma	Moderate	BRAF, NRAS

Prerequisites

• Python 3.10+
• GISTIC2 or similar for CNV analysis
• Gene signature databases
• Survival analysis packages

Related Skills

• HRD_Analysis_Agent - For HR-specific instability
• Pan_Cancer_MultiOmics_Agent - For pan-cancer context
• Tumor_Clonal_Evolution_Agent - For evolutionary dynamics

Research Applications

1. Biomarker Development: CIN as predictive marker
2. Drug Development: CIN-targeted therapy trials
3. Evolution Studies: Track CIN changes over time
4. Resistance Mechanisms: CIN and drug resistance

Author

AI Group - Biomedical AI Platform