ClonalStats Process Configuration

Purpose

Generate comprehensive clonality statistics and diversity visualizations for TCR/BCR repertoire analysis. Quantifies clonal expansion, measures diversity metrics (Shannon, Simpson, Gini), and creates publication-ready plots.

When to Use

• To quantify clonal expansion patterns in TCR/BCR data
• For diversity analysis comparing multiple samples or conditions
• To identify hyperexpanded clones and their distribution
• For rarefaction analysis to assess sampling depth
• After ScRepCombiningExpression to analyze integrated TCR+RNA data

Configuration Structure

Process Enablement

[ClonalStats]
cache = true

Input Specification

[ClonalStats.in]
screpfile = ["ScRepCombiningExpression"]

Core Environment Variables

[ClonalStats.envs]
# Clone definition: "gene" (VDJC), "aa" (CDR3 amino acid), "nt" (CDR3 nucleotide)
clone_call = "aa"
# Chain analysis: "both", "TRA", "TRB", "TRG", "IGH", "IGL"
chain = "both"
# Data transformations (dplyr::mutate syntax)
mutaters = {}
# Data filtering (dplyr::filter syntax)
subset = null
# Output device parameters
devpars = {width = 800, height = 600, res = 100}
# Save code and data (large files - use with caution)
save_code = false
save_data = false

Case-Based Plot Generation

[ClonalStats.envs.cases."Case Name"]
viz_type = "volume"  # volume, abundance, length, residency, stat,
                    # composition, overlap, diversity, geneusage,
                    # positional, kmer, rarefaction

Diversity Metrics

Metric	Range	Interpretation	Best For
shannon	0 - ∞	Higher = more diversity	General comparison
inv.simpson	1 - ∞	Higher = more diversity	Common clones
gini.coeff	0 - 1	0 = equality, 1 = inequality	Clonality dominance
norm.entropy	0 - 1	Higher = more diversity	Evenness-focused
chao1	≥ richness	Estimates total richness	Small samples
d50	Count	Clones making up 50%	Practical dominance

Interpretation:

• High diversity = Many unique clones, even distribution (healthy repertoire)
• Low diversity = Few dominant clones (antigen-specific response, infection, cancer)
• Gini ≈ 1 = Very skewed, few clones dominate
• Gini ≈ 0 = Even distribution

Visualization Types

viz_type options:

• volume - Number of clones per sample/group
• abundance - Clone abundance distribution (trend/histogram/density)
• length - CDR3 sequence length distribution
• residency - Clones present across groups (venn/upset)
• stat - Expanded clone analysis (pies/sankey)
• diversity - Diversity metrics (bar/box/violin)
• geneusage - V/D/J gene usage frequency
• rarefaction - Sampling depth assessment

Configuration Examples

Minimal Configuration

[ClonalStats.in]
screpfile = ["ScRepCombiningExpression"]

Standard Diversity Analysis

[ClonalStats.in]
screpfile = ["ScRepCombiningExpression"]

[ClonalStats.envs.cases."Diversity"]
viz_type = "diversity"
method = "shannon"
plot_type = "box"
group_by = "Diagnosis"
comparisons = true

[ClonalStats.envs.cases."Gini Coeff"]
viz_type = "diversity"
method = "gini.coeff"
plot_type = "violin"
group_by = "Diagnosis"
add_box = true

Expanded Clone Analysis

[ClonalStats.in]
screpfile = ["ScRepCombiningExpression"]

[ClonalStats.envs.cases."Expanded Clones"]
viz_type = "stat"
plot_type = "pies"
group_by = "Diagnosis"
subgroup_by = "seurat_clusters"
clones = {"Expanded (>2)" = "sel(Colitis > 2)"}

Rarefaction Analysis

[ClonalStats.in]
screpfile = ["ScRepCombiningExpression"]

[ClonalStats.envs.cases."Rarefaction"]
viz_type = "rarefaction"
group_by = "Patient"
q = 1  # 0=richness, 1=shannon, 2=simpson
n_boots = 20

Complete Analysis Suite

[ClonalStats.in]
screpfile = ["ScRepCombiningExpression"]

[ClonalStats.envs.cases."Volume"]
viz_type = "volume"

[ClonalStats.envs.cases."Abundance"]
viz_type = "abundance"
plot_type = "density"

[ClonalStats.envs.cases."Diversity"]
viz_type = "diversity"
method = "shannon"

[ClonalStats.envs.cases."Rarefaction"]
viz_type = "rarefaction"

Common Patterns

Disease vs Healthy

[ClonalStats.envs.cases."Comparison"]
viz_type = "diversity"
method = "gini.coeff"
plot_type = "box"
group_by = "Condition"
comparisons = true

Time Course

[ClonalStats.envs.cases."Timepoint"]
viz_type = "volume"
x = "Timepoint"

[ClonalStats.envs.cases."Diversity"]
viz_type = "diversity"
method = "shannon"
group_by = "Timepoint"

Treatment Response

[ClonalStats.envs.cases."Response"]
viz_type = "diversity"
method = "gini.coeff"
group_by = "Response"
plot_type = "box"
comparisons = true

Dependencies

• Upstream: ScRepCombiningExpression (required)
• Related: ScRepLoading, CDR3Clustering, TESSA (optional)

Validation Rules

• Input must be valid scRepertoire object
• For viz_type = "diversity", method must be supported
• For rarefaction, n_boots should be ≥ 10
• Use sel() syntax in clones parameter for filtering

Troubleshooting

Sample column not found: Input must have Sample column or specify x parameter.

Strange diversity values: Small repertoire sizes cause bias. Use plot_type = "box".

Rarefaction curves noisy: Increase n_boots (try 50-100).

Too many clones in stat plots: Use subset or stricter clones thresholds.

Plot generation slow: Use clone_call = "gene" for speed, apply subset.

Missing comparisons: Set comparisons = true to add significance tests.

Best Practices

1. Start with default cases to see standard visualizations
2. Use multiple diversity metrics: Shannon + Gini
3. Check rarefaction curves to ensure sufficient sampling
4. Document clone thresholds when defining expanded clones
5. Use clone_call = "gene" for speed, "aa" for granularity
6. Set save_data = true for debugging (watch disk space)
7. Validate findings with complementary diversity indices
8. Consider sample size: small samples underestimate richness