@bgicli/bgicli 2.1.1 → 2.2.1

package/data/skills/clinvar-database/SKILL.md

@@ -0,0 +1,356 @@
+---
+name: clinvar-database
+description: "Query NCBI ClinVar for variant clinical significance. Search by gene/position, interpret pathogenicity classifications, access via E-utilities API or FTP, annotate VCFs, for genomic medicine."
+---
+
+# ClinVar Database
+
+## Overview
+
+ClinVar is NCBI's freely accessible archive of reports on relationships between human genetic variants and phenotypes, with supporting evidence. The database aggregates information about genomic variation and its relationship to human health, providing standardized variant classifications used in clinical genetics and research.
+
+## When to Use This Skill
+
+This skill should be used when:
+
+- Searching for variants by gene, condition, or clinical significance
+- Interpreting clinical significance classifications (pathogenic, benign, VUS)
+- Accessing ClinVar data programmatically via E-utilities API
+- Downloading and processing bulk data from FTP
+- Understanding review status and star ratings
+- Resolving conflicting variant interpretations
+- Annotating variant call sets with clinical significance
+
+## Core Capabilities
+
+### 1. Search and Query ClinVar
+
+#### Web Interface Queries
+
+Search ClinVar using the web interface at https://www.ncbi.nlm.nih.gov/clinvar/
+
+**Common search patterns:**
+- By gene: `BRCA1[gene]`
+- By clinical significance: `pathogenic[CLNSIG]`
+- By condition: `breast cancer[disorder]`
+- By variant: `NM_000059.3:c.1310_1313del[variant name]`
+- By chromosome: `13[chr]`
+- Combined: `BRCA1[gene] AND pathogenic[CLNSIG]`
+
+#### Programmatic Access via E-utilities
+
+Access ClinVar programmatically using NCBI's E-utilities API. Refer to `references/api_reference.md` for comprehensive API documentation including:
+- **esearch** - Search for variants matching criteria
+- **esummary** - Retrieve variant summaries
+- **efetch** - Download full XML records
+- **elink** - Find related records in other NCBI databases
+
+**Quick example using curl:**
+```bash
+# Search for pathogenic BRCA1 variants
+curl "https://eutils.ncbi.nlm.nih.gov/entrez/eutils/esearch.fcgi?db=clinvar&term=BRCA1[gene]+AND+pathogenic[CLNSIG]&retmode=json"
+```
+
+**Best practices:**
+- Test queries on the web interface before automating
+- Use API keys to increase rate limits from 3 to 10 requests/second
+- Implement exponential backoff for rate limit errors
+- Set `Entrez.email` when using Biopython
+
+### 2. Interpret Clinical Significance
+
+#### Understanding Classifications
+
+ClinVar uses standardized terminology for variant classifications. Refer to `references/clinical_significance.md` for detailed interpretation guidelines.
+
+**Key germline classification terms (ACMG/AMP):**
+- **Pathogenic (P)** - Variant causes disease (~99% probability)
+- **Likely Pathogenic (LP)** - Variant likely causes disease (~90% probability)
+- **Uncertain Significance (VUS)** - Insufficient evidence to classify
+- **Likely Benign (LB)** - Variant likely does not cause disease
+- **Benign (B)** - Variant does not cause disease
+
+**Review status (star ratings):**
+- ★★★★ Practice guideline - Highest confidence
+- ★★★ Expert panel review (e.g., ClinGen) - High confidence
+- ★★ Multiple submitters, no conflicts - Moderate confidence
+- ★ Single submitter with criteria - Standard weight
+- ☆ No assertion criteria - Low confidence
+
+**Critical considerations:**
+- Always check review status - prefer ★★★ or ★★★★ ratings
+- Conflicting interpretations require manual evaluation
+- Classifications may change as new evidence emerges
+- VUS (uncertain significance) variants lack sufficient evidence for clinical use
+
+### 3. Download Bulk Data from FTP
+
+#### Access ClinVar FTP Site
+
+Download complete datasets from `ftp://ftp.ncbi.nlm.nih.gov/pub/clinvar/`
+
+Refer to `references/data_formats.md` for comprehensive documentation on file formats and processing.
+
+**Update schedule:**
+- Monthly releases: First Thursday of each month (complete dataset, archived)
+- Weekly updates: Every Monday (incremental updates)
+
+#### Available Formats
+
+**XML files** (most comprehensive):
+- VCV (Variation) files: `xml/clinvar_variation/` - Variant-centric aggregation
+- RCV (Record) files: `xml/RCV/` - Variant-condition pairs
+- Include full submission details, evidence, and metadata
+
+**VCF files** (for genomic pipelines):
+- GRCh37: `vcf_GRCh37/clinvar.vcf.gz`
+- GRCh38: `vcf_GRCh38/clinvar.vcf.gz`
+- Limitations: Excludes variants >10kb and complex structural variants
+
+**Tab-delimited files** (for quick analysis):
+- `tab_delimited/variant_summary.txt.gz` - Summary of all variants
+- `tab_delimited/var_citations.txt.gz` - PubMed citations
+- `tab_delimited/cross_references.txt.gz` - Database cross-references
+
+**Example download:**
+```bash
+# Download latest monthly XML release
+wget ftp://ftp.ncbi.nlm.nih.gov/pub/clinvar/xml/clinvar_variation/ClinVarVariationRelease_00-latest.xml.gz
+
+# Download VCF for GRCh38
+wget ftp://ftp.ncbi.nlm.nih.gov/pub/clinvar/vcf_GRCh38/clinvar.vcf.gz
+```
+
+### 4. Process and Analyze ClinVar Data
+
+#### Working with XML Files
+
+Process XML files to extract variant details, classifications, and evidence.
+
+**Python example with xml.etree:**
+```python
+import gzip
+import xml.etree.ElementTree as ET
+
+with gzip.open('ClinVarVariationRelease.xml.gz', 'rt') as f:
+    for event, elem in ET.iterparse(f, events=('end',)):
+        if elem.tag == 'VariationArchive':
+            variation_id = elem.attrib.get('VariationID')
+            # Extract clinical significance, review status, etc.
+            elem.clear()  # Free memory
+```
+
+#### Working with VCF Files
+
+Annotate variant calls or filter by clinical significance using bcftools or Python.
+
+**Using bcftools:**
+```bash
+# Filter pathogenic variants
+bcftools view -i 'INFO/CLNSIG~"Pathogenic"' clinvar.vcf.gz
+
+# Extract specific genes
+bcftools view -i 'INFO/GENEINFO~"BRCA"' clinvar.vcf.gz
+
+# Annotate your VCF with ClinVar
+bcftools annotate -a clinvar.vcf.gz -c INFO your_variants.vcf
+```
+
+**Using PyVCF in Python:**
+```python
+import vcf
+
+vcf_reader = vcf.Reader(filename='clinvar.vcf.gz')
+for record in vcf_reader:
+    clnsig = record.INFO.get('CLNSIG', [])
+    if 'Pathogenic' in clnsig:
+        gene = record.INFO.get('GENEINFO', [''])[0]
+        print(f"{record.CHROM}:{record.POS} {gene} - {clnsig}")
+```
+
+#### Working with Tab-Delimited Files
+
+Use pandas or command-line tools for rapid filtering and analysis.
+
+**Using pandas:**
+```python
+import pandas as pd
+
+# Load variant summary
+df = pd.read_csv('variant_summary.txt.gz', sep='\t', compression='gzip')
+
+# Filter pathogenic variants in specific gene
+pathogenic_brca = df[
+    (df['GeneSymbol'] == 'BRCA1') &
+    (df['ClinicalSignificance'].str.contains('Pathogenic', na=False))
+]
+
+# Count variants by clinical significance
+sig_counts = df['ClinicalSignificance'].value_counts()
+```
+
+**Using command-line tools:**
+```bash
+# Extract pathogenic variants for specific gene
+zcat variant_summary.txt.gz | \
+  awk -F'\t' '$7=="TP53" && $13~"Pathogenic"' | \
+  cut -f1,5,7,13,14
+```
+
+### 5. Handle Conflicting Interpretations
+
+When multiple submitters provide different classifications for the same variant, ClinVar reports "Conflicting interpretations of pathogenicity."
+
+**Resolution strategy:**
+1. Check review status (star rating) - higher ratings carry more weight
+2. Examine evidence and assertion criteria from each submitter
+3. Consider submission dates - newer submissions may reflect updated evidence
+4. Review population frequency data (e.g., gnomAD) for context
+5. Consult expert panel classifications (★★★) when available
+6. For clinical use, always defer to a genetics professional
+
+**Search query to exclude conflicts:**
+```
+TP53[gene] AND pathogenic[CLNSIG] NOT conflicting[RVSTAT]
+```
+
+### 6. Track Classification Updates
+
+Variant classifications may change over time as new evidence emerges.
+
+**Why classifications change:**
+- New functional studies or clinical data
+- Updated population frequency information
+- Revised ACMG/AMP guidelines
+- Segregation data from additional families
+
+**Best practices:**
+- Document ClinVar version and access date for reproducibility
+- Re-check classifications periodically for critical variants
+- Subscribe to ClinVar mailing list for major updates
+- Use monthly archived releases for stable datasets
+
+### 7. Submit Data to ClinVar
+
+Organizations can submit variant interpretations to ClinVar.
+
+**Submission methods:**
+- Web submission portal: https://submit.ncbi.nlm.nih.gov/subs/clinvar/
+- API submission (requires service account): See `references/api_reference.md`
+- Batch submission via Excel templates
+
+**Requirements:**
+- Organizational account with NCBI
+- Assertion criteria (preferably ACMG/AMP guidelines)
+- Supporting evidence for classification
+
+Contact: clinvar@ncbi.nlm.nih.gov for submission account setup.
+
+## Workflow Examples
+
+### Example 1: Identify High-Confidence Pathogenic Variants in a Gene
+
+**Objective:** Find pathogenic variants in CFTR gene with expert panel review.
+
+**Steps:**
+1. Search using web interface or E-utilities:
+   ```
+   CFTR[gene] AND pathogenic[CLNSIG] AND (reviewed by expert panel[RVSTAT] OR practice guideline[RVSTAT])
+   ```
+2. Review results, noting review status (should be ★★★ or ★★★★)
+3. Export variant list or retrieve full records via efetch
+4. Cross-reference with clinical presentation if applicable
+
+### Example 2: Annotate VCF with ClinVar Classifications
+
+**Objective:** Add clinical significance annotations to variant calls.
+
+**Steps:**
+1. Download appropriate ClinVar VCF (match genome build: GRCh37 or GRCh38):
+   ```bash
+   wget ftp://ftp.ncbi.nlm.nih.gov/pub/clinvar/vcf_GRCh38/clinvar.vcf.gz
+   wget ftp://ftp.ncbi.nlm.nih.gov/pub/clinvar/vcf_GRCh38/clinvar.vcf.gz.tbi
+   ```
+2. Annotate using bcftools:
+   ```bash
+   bcftools annotate -a clinvar.vcf.gz \
+     -c INFO/CLNSIG,INFO/CLNDN,INFO/CLNREVSTAT \
+     -o annotated_variants.vcf \
+     your_variants.vcf
+   ```
+3. Filter annotated VCF for pathogenic variants:
+   ```bash
+   bcftools view -i 'INFO/CLNSIG~"Pathogenic"' annotated_variants.vcf
+   ```
+
+### Example 3: Analyze Variants for a Specific Disease
+
+**Objective:** Study all variants associated with hereditary breast cancer.
+
+**Steps:**
+1. Search by condition:
+   ```
+   hereditary breast cancer[disorder] OR "Breast-ovarian cancer, familial"[disorder]
+   ```
+2. Download results as CSV or retrieve via E-utilities
+3. Filter by review status to prioritize high-confidence variants
+4. Analyze distribution across genes (BRCA1, BRCA2, PALB2, etc.)
+5. Examine variants with conflicting interpretations separately
+
+### Example 4: Bulk Download and Database Construction
+
+**Objective:** Build a local ClinVar database for analysis pipeline.
+
+**Steps:**
+1. Download monthly release for reproducibility:
+   ```bash
+   wget ftp://ftp.ncbi.nlm.nih.gov/pub/clinvar/xml/clinvar_variation/ClinVarVariationRelease_YYYY-MM.xml.gz
+   ```
+2. Parse XML and load into database (PostgreSQL, MySQL, MongoDB)
+3. Index by gene, position, clinical significance, review status
+4. Implement version tracking for updates
+5. Schedule monthly updates from FTP site
+
+## Important Limitations and Considerations
+
+### Data Quality
+- **Not all submissions have equal weight** - Check review status (star ratings)
+- **Conflicting interpretations exist** - Require manual evaluation
+- **Historical submissions may be outdated** - Newer data may be more accurate
+- **VUS classification is not a clinical diagnosis** - Means insufficient evidence
+
+### Scope Limitations
+- **Not for direct clinical diagnosis** - Always involve genetics professional
+- **Population-specific** - Variant frequencies vary by ancestry
+- **Incomplete coverage** - Not all genes or variants are well-studied
+- **Version dependencies** - Coordinate genome build (GRCh37/GRCh38) across analyses
+
+### Technical Limitations
+- **VCF files exclude large variants** - Variants >10kb not in VCF format
+- **Rate limits on API** - 3 req/sec without key, 10 req/sec with API key
+- **File sizes** - Full XML releases are multi-GB compressed files
+- **No real-time updates** - Website updated weekly, FTP monthly/weekly
+
+## Resources
+
+### Reference Documentation
+
+This skill includes comprehensive reference documentation:
+
+- **`references/api_reference.md`** - Complete E-utilities API documentation with examples for esearch, esummary, efetch, and elink; includes rate limits, authentication, and Python/Biopython code samples
+
+- **`references/clinical_significance.md`** - Detailed guide to interpreting clinical significance classifications, review status star ratings, conflict resolution, and best practices for variant interpretation
+
+- **`references/data_formats.md`** - Documentation for XML, VCF, and tab-delimited file formats; FTP directory structure, processing examples, and format selection guidance
+
+### External Resources
+
+- ClinVar home: https://www.ncbi.nlm.nih.gov/clinvar/
+- ClinVar documentation: https://www.ncbi.nlm.nih.gov/clinvar/docs/
+- E-utilities documentation: https://www.ncbi.nlm.nih.gov/books/NBK25501/
+- ACMG variant interpretation guidelines: Richards et al., 2015 (PMID: 25741868)
+- ClinGen expert panels: https://clinicalgenome.org/
+
+### Contact
+
+For questions about ClinVar or data submission: clinvar@ncbi.nlm.nih.gov

package/data/skills/cnv-caller-agent/SKILL.md

@@ -0,0 +1,171 @@
+<!--
+# COPYRIGHT NOTICE
+# This file is part of the "Universal Biomedical Skills" project.
+# Copyright (c) 2026 MD BABU MIA, PhD <md.babu.mia@mssm.edu>
+# All Rights Reserved.
+#
+# This code is proprietary and confidential.
+# Unauthorized copying of this file, via any medium is strictly prohibited.
+#
+# Provenance: Authenticated by MD BABU MIA
+
+-->
+
+---
+name: 'cnv-caller-agent'
+description: 'AI-enhanced copy number variation calling and analysis from sequencing data for cancer genomics, constitutional CNV detection, and chromosomal aberration characterization.'
+measurable_outcome: Execute skill workflow successfully with valid output within 15 minutes.
+allowed-tools:
+  - read_file
+  - run_shell_command
+---
+
+
+# CNV Caller Agent
+
+The **CNV Caller Agent** provides comprehensive AI-enhanced copy number variation analysis from WGS, WES, and targeted sequencing for cancer genomics and constitutional CNV detection.
+
+## When to Use This Skill
+
+* When calling somatic CNVs from tumor-normal paired sequencing.
+* To detect constitutional CNVs from germline sequencing.
+* For allele-specific copy number analysis.
+* When characterizing focal amplifications and deletions in cancer.
+* To assess tumor purity and ploidy from CNV data.
+
+## Core Capabilities
+
+1. **Somatic CNV Calling**: Detect tumor-specific copy number alterations.
+
+2. **Germline CNV Detection**: Identify constitutional CNVs for rare disease.
+
+3. **Allele-Specific Analysis**: Determine allele-specific copy number and LOH.
+
+4. **Purity/Ploidy Estimation**: Estimate tumor content and genome doubling.
+
+5. **Focal Event Detection**: Identify amplifications and deletions of driver genes.
+
+6. **Segmentation Optimization**: AI-enhanced breakpoint detection.
+
+## Workflow
+
+1. **Input**: BAM files (tumor/normal), or targeted panel data.
+
+2. **Coverage Normalization**: GC correction, mappability adjustment.
+
+3. **Segmentation**: Identify regions of consistent copy number.
+
+4. **Allele-Specific**: Calculate B-allele frequency for heterozygosity.
+
+5. **Purity/Ploidy**: Estimate sample parameters.
+
+6. **Calling**: Assign integer copy number states.
+
+7. **Output**: Segmented CNV calls, purity/ploidy, driver events.
+
+## Example Usage
+
+**User**: "Call somatic copy number alterations from this tumor-normal WES pair."
+
+**Agent Action**:
+```bash
+python3 Skills/Genomics/CNV_Caller_Agent/cnv_caller.py \
+    --tumor tumor.bam \
+    --normal normal.bam \
+    --reference GRCh38.fa \
+    --method facets \
+    --targets exome_targets.bed \
+    --driver_genes cancer_genes.txt \
+    --output cnv_results/
+```
+
+## CNV Calling Methods
+
+| Tool | Application | Key Features |
+|------|-------------|--------------|
+| FACETS | Tumor WES | Purity/ploidy, allele-specific |
+| ASCAT | Tumor WGS/arrays | Allele-specific, multi-clone |
+| CNVkit | WES/targeted | Hybrid reference approach |
+| GATK CNV | WES/WGS | GATK ecosystem integration |
+| Purple | WGS | GRIDSS integration, comprehensive |
+| CONICS | scRNA-seq | Single-cell CNV inference |
+
+## Key Output Metrics
+
+| Metric | Description | Interpretation |
+|--------|-------------|----------------|
+| Purity | Tumor fraction | Sample quality |
+| Ploidy | Average copy number | Genome doubling |
+| LOH | Loss of heterozygosity | Regions of allele loss |
+| SCNA burden | Total altered fraction | Genomic instability |
+| Focal events | Amplifications/deletions | Driver candidates |
+
+## Cancer Driver CNVs
+
+| Gene | Alteration | Cancer Type |
+|------|------------|-------------|
+| ERBB2 (HER2) | Amplification | Breast, gastric |
+| MYC | Amplification | Many cancers |
+| EGFR | Amplification | Lung, GBM |
+| CDK4/MDM2 | Amplification | Sarcoma, GBM |
+| CDKN2A | Deletion | Many cancers |
+| RB1 | Deletion | Many cancers |
+| PTEN | Deletion | Prostate, GBM |
+
+## AI/ML Enhancements
+
+**Segmentation**:
+- Deep learning for breakpoint detection
+- Noise reduction in low-coverage data
+- Improved sensitivity for focal events
+
+**Quality Prediction**:
+- Sample quality scoring
+- Artifact detection
+- Confidence estimation
+
+**Driver Prioritization**:
+- GISTIC-style analysis
+- Functional impact scoring
+- Pan-cancer frequency context
+
+## Allele-Specific Copy Number
+
+```
+Total CN = Major allele + Minor allele
+
+Examples:
+- Normal: 1 + 1 = 2 (diploid)
+- CN gain: 2 + 1 = 3 (trisomy)
+- CN-LOH: 2 + 0 = 2 (normal total, LOH)
+- Homozygous deletion: 0 + 0 = 0
+- High amplification: 10 + 0 = 10 (focal amp)
+```
+
+## Prerequisites
+
+* Python 3.10+
+* CNV calling tools (FACETS, CNVkit, etc.)
+* Reference genome and annotations
+* Sufficient memory for WGS (16GB+)
+
+## Related Skills
+
+* Variant_Interpretation - For CNV annotation
+* HRD_Analysis_Agent - For HRD scoring from CNV
+* Pan_Cancer_MultiOmics_Agent - For pan-cancer CNV context
+
+## Quality Considerations
+
+1. **Coverage depth**: Higher = better resolution
+2. **Tumor purity**: Low purity challenges calling
+3. **Normal match**: Best with matched normal
+4. **Target design**: Uniform coverage for panels
+5. **GC bias**: Proper normalization critical
+
+## Author
+
+AI Group - Biomedical AI Platform
+
+
+<!-- AUTHOR_SIGNATURE: 9a7f3c2e-MD-BABU-MIA-2026-MSSM-SECURE -->

package/data/skills/coagulation-thrombosis-agent/SKILL.md

@@ -0,0 +1,141 @@
+<!--
+# COPYRIGHT NOTICE
+# This file is part of the "Universal Biomedical Skills" project.
+# Copyright (c) 2026 MD BABU MIA, PhD <md.babu.mia@mssm.edu>
+# All Rights Reserved.
+#
+# This code is proprietary and confidential.
+# Unauthorized copying of this file, via any medium is strictly prohibited.
+#
+# Provenance: Authenticated by MD BABU MIA
+
+-->
+
+---
+name: 'coagulation-thrombosis-agent'
+description: 'AI-powered analysis of coagulation disorders, thrombosis risk prediction, anticoagulation management, and platelet function assessment using machine learning.'
+measurable_outcome: Execute skill workflow successfully with valid output within 15 minutes.
+allowed-tools:
+  - read_file
+  - run_shell_command
+---
+
+
+# Coagulation and Thrombosis Agent
+
+The **Coagulation and Thrombosis Agent** provides AI-driven analysis of hemostatic disorders, thrombosis risk assessment, and anticoagulation management. It integrates coagulation cascade modeling, platelet function analysis, and machine learning for personalized thrombosis prevention.
+
+## When to Use This Skill
+
+* When assessing venous thromboembolism (VTE) risk in hospitalized patients.
+* For anticoagulation dose optimization (warfarin, DOACs).
+* To analyze coagulation panel results and identify bleeding/clotting disorders.
+* For platelet morphology and function assessment.
+* When managing thrombosis in myeloproliferative neoplasms (MPNs).
+
+## Core Capabilities
+
+1. **VTE Risk Prediction**: Machine learning models predict deep vein thrombosis (DVT) and pulmonary embolism (PE) risk using clinical and laboratory features.
+
+2. **Anticoagulation Optimization**: AI-guided dosing for warfarin (incorporating pharmacogenomics) and monitoring for DOACs.
+
+3. **Coagulation Cascade Analysis**: Interprets PT, aPTT, fibrinogen, D-dimer, and specialized assays to diagnose coagulopathies.
+
+4. **Platelet Analysis**: CNN-based morphology analysis predicting bleeding and thrombosis risk from peripheral smear images.
+
+5. **DIC Scoring**: Automated disseminated intravascular coagulation (DIC) scoring and monitoring.
+
+6. **MPN Thrombosis Risk**: Specialized models for thrombosis prediction in polycythemia vera, essential thrombocythemia.
+
+## Workflow
+
+1. **Input**: Coagulation lab results, patient demographics, clinical risk factors, platelet images (optional).
+
+2. **Risk Assessment**: Apply ML models for VTE, bleeding, or DIC risk scores.
+
+3. **Dosing Optimization**: Generate anticoagulation recommendations.
+
+4. **Monitoring**: Track INR/anti-Xa trends and alert on deviations.
+
+5. **Diagnosis**: Pattern recognition for coagulation disorders.
+
+6. **Output**: Risk scores, dosing recommendations, diagnostic suggestions, monitoring alerts.
+
+## Example Usage
+
+**User**: "Calculate VTE risk for this hospitalized patient and optimize LMWH prophylaxis."
+
+**Agent Action**:
+```bash
+python3 Skills/Hematology/Coagulation_Thrombosis_Agent/thrombosis_analyzer.py \
+    --patient_data patient_demographics.json \
+    --labs coagulation_panel.csv \
+    --risk_model improved_padua \
+    --anticoagulant lmwh \
+    --renal_function egfr_45 \
+    --output vte_assessment.json
+```
+
+## Risk Models Implemented
+
+| Model | Application | Key Features |
+|-------|-------------|--------------|
+| Padua (Enhanced) | Medical VTE risk | 11 clinical factors + ML enhancement |
+| Caprini (AI) | Surgical VTE risk | 40+ factors with ML weighting |
+| CHADS2-VASc | Atrial fibrillation stroke risk | Standard guideline scoring |
+| HAS-BLED | Anticoagulation bleeding risk | Major bleeding prediction |
+| IPSET-thrombosis | MPN thrombosis | JAK2, age, prior thrombosis |
+
+## Coagulation Panel Interpretation
+
+| Test | Normal Range | Elevations Suggest | Decreases Suggest |
+|------|--------------|-------------------|-------------------|
+| PT/INR | 11-13.5s / 0.9-1.1 | Warfarin, VII def, liver disease | - |
+| aPTT | 25-35s | Heparin, VIII/IX/XI def, lupus AC | - |
+| Fibrinogen | 200-400 mg/dL | Acute phase, inflammation | DIC, liver disease |
+| D-dimer | <500 ng/mL | VTE, DIC, inflammation | - |
+| Platelet | 150-400K | Reactive, MPN | ITP, marrow failure |
+
+## AI/ML Components
+
+**Deep Learning for Platelet Morphology**:
+- CNN analysis of peripheral smear images
+- Identifies giant platelets, platelet clumps, hypogranular forms
+- Predicts bleeding/thrombosis risk from morphology
+
+**VTE Prediction Models**:
+- Gradient boosting (XGBoost) on structured EHR data
+- Incorporates labs, vitals, medications, procedures
+- AUC > 0.85 for hospital-acquired VTE
+
+**Anticoagulation Dosing**:
+- Reinforcement learning for INR control
+- Pharmacogenomic integration (CYP2C9, VKORC1)
+- Real-time dose adjustment recommendations
+
+## Prerequisites
+
+* Python 3.10+
+* scikit-learn, XGBoost, PyTorch
+* HL7 FHIR client (for EHR integration)
+* Image analysis libraries (for platelet morphology)
+
+## Related Skills
+
+* Flow_Cytometry_AI - For platelet function assays
+* Pharmacogenomics_Agent - For anticoagulant pharmacogenomics
+* Blood_Smear_Analysis - For morphology assessment
+
+## Clinical Applications
+
+1. **Hospital VTE Prevention**: Real-time risk scoring in EMR
+2. **Anticoagulation Clinic**: AI-assisted warfarin dosing
+3. **DIC Management**: Automated scoring and transfusion guidance
+4. **Inherited Disorders**: Pattern recognition for factor deficiencies
+
+## Author
+
+AI Group - Biomedical AI Platform
+
+
+<!-- AUTHOR_SIGNATURE: 9a7f3c2e-MD-BABU-MIA-2026-MSSM-SECURE -->