Variant Surveillance

<!--
# COPYRIGHT NOTICE
# This file is part of the "Universal AI Agentic 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: bio-epidemiological-genomics-variant-surveillance
description: Assign pathogen lineages and track variants using Nextclade and pangolin for viral surveillance. Monitor variant prevalence and identify emerging variants of concern. Use when classifying viral sequences, tracking lineage dynamics, or monitoring for variants of concern.
tool_type: cli
primary_tool: nextclade
measurable_outcome: Execute skill workflow successfully with valid output within 15 minutes.
allowed-tools:
  - read_file
  - run_shell_command
---

# Variant Surveillance

## Nextclade CLI

```bash
# Install Nextclade
npm install -g @nextstrain/nextclade

# Or download binary
curl -fsSL "https://github.com/nextstrain/nextclade/releases/latest/download/nextclade-x86_64-unknown-linux-gnu" -o nextclade
chmod +x nextclade

# List available datasets
nextclade dataset list

# Download dataset (e.g., SARS-CoV-2)
nextclade dataset get --name sars-cov-2 --output-dir data/sars-cov-2

# Run analysis
nextclade run \
    --input-dataset data/sars-cov-2 \
    --output-tsv results.tsv \
    --output-json results.json \
    sequences.fasta
```

## Pangolin for SARS-CoV-2

```bash
# Install pangolin
pip install pangolin

# Update lineage definitions
pangolin --update

# Run lineage assignment
pangolin sequences.fasta -o pangolin_results.csv

# With specific version
pangolin sequences.fasta --analysis-mode accurate -o results.csv
```

## Parse Nextclade Results

```python
import pandas as pd

def parse_nextclade(results_file):
    '''Parse Nextclade TSV output

    Key columns:
    - seqName: Sequence identifier
    - clade: Nextstrain clade (e.g., 21L for Omicron BA.2)
    - Nextclade_pango: Pangolin lineage
    - qc.overallStatus: Quality control status
    - substitutions: List of mutations
    - aaSubstitutions: Amino acid changes
    '''
    df = pd.read_csv(results_file, sep='\t')

    # Filter by QC status
    df['pass_qc'] = df['qc.overallStatus'].isin(['good', 'mediocre'])

    return df


def summarize_lineages(results_df, lineage_col='Nextclade_pango'):
    '''Summarize lineage distribution'''
    # Filter passed QC
    passed = results_df[results_df['pass_qc']]

    summary = {
        'total_sequences': len(results_df),
        'passed_qc': len(passed),
        'unique_lineages': passed[lineage_col].nunique(),
        'lineage_counts': passed[lineage_col].value_counts().to_dict()
    }

    return summary
```

## Track Variants of Concern

```python
# WHO Variants of Concern/Interest definitions
VOC_DEFINITIONS = {
    'Alpha': {'lineages': ['B.1.1.7', 'Q.*'], 'key_mutations': ['N501Y', 'P681H']},
    'Beta': {'lineages': ['B.1.351'], 'key_mutations': ['K417N', 'E484K', 'N501Y']},
    'Gamma': {'lineages': ['P.1'], 'key_mutations': ['K417T', 'E484K', 'N501Y']},
    'Delta': {'lineages': ['B.1.617.2', 'AY.*'], 'key_mutations': ['L452R', 'P681R']},
    'Omicron': {'lineages': ['B.1.1.529', 'BA.*', 'XBB.*'], 'key_mutations': ['G339D', 'N501Y']}
}

def classify_voc(lineage):
    '''Classify lineage as VOC'''
    for voc, definition in VOC_DEFINITIONS.items():
        for pattern in definition['lineages']:
            if pattern.endswith('*'):
                if lineage.startswith(pattern[:-1]):
                    return voc
            elif lineage == pattern:
                return voc
    return 'Other'


def track_voc_prevalence(results_df, date_col='collection_date'):
    '''Track VOC prevalence over time'''
    results_df = results_df.copy()
    results_df['VOC'] = results_df['Nextclade_pango'].apply(classify_voc)

    # Group by week
    results_df['week'] = pd.to_datetime(results_df[date_col]).dt.to_period('W')

    prevalence = results_df.groupby(['week', 'VOC']).size().unstack(fill_value=0)
    prevalence_pct = prevalence.div(prevalence.sum(axis=1), axis=0) * 100

    return prevalence_pct
```

## Mutation Analysis

```python
def parse_mutations(mutation_string):
    '''Parse Nextclade mutation string

    Format: 'A123T,C456G' (nucleotide) or 'S:N501Y,S:D614G' (amino acid)
    '''
    if pd.isna(mutation_string) or mutation_string == '':
        return []
    return mutation_string.split(',')


def find_mutation_prevalence(results_df, mutation_col='aaSubstitutions'):
    '''Calculate prevalence of each mutation'''
    all_mutations = []
    for muts in results_df[mutation_col].dropna():
        all_mutations.extend(parse_mutations(muts))

    mutation_counts = pd.Series(all_mutations).value_counts()
    mutation_prevalence = mutation_counts / len(results_df) * 100

    return mutation_prevalence


def detect_emerging_mutations(results_df, date_col='collection_date', threshold=5):
    '''Detect mutations increasing in frequency

    Alerts for mutations that:
    1. Were rare in early period
    2. Increased significantly in recent period
    '''
    results_df = results_df.copy()
    results_df['date'] = pd.to_datetime(results_df[date_col])

    # Split into early and recent
    midpoint = results_df['date'].median()
    early = results_df[results_df['date'] < midpoint]
    recent = results_df[results_df['date'] >= midpoint]

    early_prev = find_mutation_prevalence(early)
    recent_prev = find_mutation_prevalence(recent)

    # Find emerging (low->high)
    emerging = []
    for mut in recent_prev.index:
        early_freq = early_prev.get(mut, 0)
        recent_freq = recent_prev[mut]

        if recent_freq > threshold and recent_freq > early_freq * 2:
            emerging.append({
                'mutation': mut,
                'early_prevalence': early_freq,
                'recent_prevalence': recent_freq,
                'fold_change': recent_freq / max(early_freq, 0.1)
            })

    return sorted(emerging, key=lambda x: -x['fold_change'])
```

## Surveillance Report

```python
def generate_surveillance_report(results_df, period='week'):
    '''Generate variant surveillance report'''
    passed = results_df[results_df['pass_qc']]

    report = {
        'period': period,
        'total_sequences': len(results_df),
        'passed_qc': len(passed),
        'qc_pass_rate': f"{len(passed)/len(results_df)*100:.1f}%"
    }

    # Lineage distribution
    lineage_counts = passed['Nextclade_pango'].value_counts()
    report['dominant_lineage'] = lineage_counts.index[0]
    report['dominant_lineage_pct'] = f"{lineage_counts.iloc[0]/len(passed)*100:.1f}%"
    report['top_5_lineages'] = lineage_counts.head(5).to_dict()

    # VOC tracking
    passed['VOC'] = passed['Nextclade_pango'].apply(classify_voc)
    voc_counts = passed['VOC'].value_counts()
    report['voc_distribution'] = voc_counts.to_dict()

    return report
```

## Related Skills

- epidemiological-genomics/phylodynamics - Time-scaled analysis of variants
- variant-calling/variant-annotation - Mutation annotation
- data-visualization/ggplot2-fundamentals - Visualize variant dynamics


<!-- AUTHOR_SIGNATURE: 9a7f3c2e-MD-BABU-MIA-2026-MSSM-SECURE -->