Entropy Clustering, Hellinger Distance, and Change Point Analysis for Emerging Viral Variant Detection: A Simulation Study | 1 Introduction | 2 Setup | 3 Reference Sequence | 4 Simulation via simulate_variant_evolution() | 4.1 Simulate Data | 4.2 Check the Pool | 4.3 Emergence Periods and Sites | 4.4 Deleterious Periods and Sites | 4.5 Inspecting the Simulation Output | 4.6 Reference Phase 1 (month 1) | 4.7 Reference Phase 4 (month 4) | 4.8 Variant 1 Phase 1 (month 5): sites 12, 15 | 4.9 Variant 1 Phase 2 (month 6): sites 12, 15 | 4.10 Variant 1 Phase 4 (month 8): sites 12, 15 | 4.11 Variant 2 Phase 1 (month 9): sites 6, 13, 23 | 4.12 Variant 2 Phase 5 (month 13): sites 6, 13, 23 | 4.13 Variant 3 Phase 1 (month 14): sites 5, 14, 16, 24 | 5 Sites Detection and Change Point Detection | 5.1 Encoding | 5.2 Partitioning | 5.3 Hellinger Distance (HD) | 5.4 Change Point Detection | 6 Conclusion | References

NCBI SARS-CoV-2 Spike Protein Sequence Preprocessing: From Raw FASTA to an Analysis-Ready Integer-Encoded Matrix | 1 Introduction | 2 Setup | 3 Step 1: Load the FASTA Dataset | 3.1 Download Data | 3.2 Update Path | 3.3 Read FASTA file | 3.4 Curated metadata for 12 SARS-CoV-2 VOCs and VOIs | 3.5 Bundled FASTA sample | 4 Step 2: Canonical Spike Protein Length | 5 Step 3: Extract and Validate Collection Dates | 5.1 Initial Date Extraction (yyyy-mm-dd) | 5.2 Second Pass: yyyy-mm Extraction | 5.3 Date Distribution (Before Ambiguous Residue Filtering) | 6 Step 4: Extract and Validate Countries | 7 Step 5: Convert, Filter, and Encode the Sequence Matrix | 7.1 Convert FASTA to Character Matrix | 7.2 Filter Ambiguous Residues | 7.3 Encode Amino Acids to Integers | 8 Step 6: Assemble the Final Data Frame | 8.1 Date Distribution (After Filtering) | 8.2 Filter US Sequences and Save | 9 Downstream Analysis | 10 Pipeline Summary | References

Unsupervised Recovery of SARS-CoV-2 Variant Structure via Entropy-Driven Site Selection and PAM Clustering: Precision, Recall, and F1 Evaluation Across Wild-Type and Delta-Dominated Surveillance Periods | 1 Introduction | 1.1 Definitions | 1.2 Accuracy Study Logic | 2 Libraries | 3 Global Parameters | 4 Load Preprocessed Data | 4.1 Country Filter | 5 Step 1: Data Partitioning | 5.1 Partition into Two Time Windows (Wild-Type vs Delta) | 5.2 Per-Period Entropy Clustering | 6 Step 2: Combine Both Partitions | 6.1 Analyze the Selected Sites | 6.2 Cross-Reference with Known Variant Sites | 7 Step 3: PAM Clustering with Gower Distance | 7.1 Prepare Feature Matrix | 7.2 Gower Distance Matrix | 7.3 Silhouette Analysis | 8 Step 4: PAM Fitting and t-SNE Visualisation | 8.1 t-SNE Embedding | 8.2 VOC SNP Matching Helpers | 8.3 t-SNE Plot Function | 8.4 PAM and Medoid Analysis | 8.5 PAM k = 2 | 8.6 PAM k = 3 | 8.7 PAM k = 4 | 8.8 PAM k = 9 | 9 Step 5: Modal Amino Acid Profile per Cluster | 10 Step 6: Precision, Recall and F1 Score | 11 Step 7: Site Differentiation — Wild-Type Cluster vs Later Clusters | 11.1 Annotation helper | 11.2 Contrasts calculation | 11.3 Biological Interpretation of Key Differentiating Sites | 12 Step 8: Borderline Period 1 Sequences | 13 Step 9: Full-Dataset Clustering (All 109,536 Sequences) | 13.1 Site Selection on Full Dataset | 13.2 Gower Distance | 13.3 Silhouette Analysis | 13.4 PAM Fitting | 13.5 t-SNE Visualisation | 13.6 Full Dataset PAM k = 6 | 13.7 Full Dataset PAM k = 7 | 13.8 Full Dataset PAM k = 9 | 14 Summary (Clustering Accuracy) | 15 Conclusions | 15.1 Clustering Accuracy Study (Steps 1–8) | 15.2 Full-Dataset Clustering: Global Separation Patterns (Step 9) | References

LIBRARIES | EXAMPLES | 1. LM object | 1.1 Data | 1.2 Default side-by-side plots | 1.3 Visualizing per-unit change together with an intercept | 1.4 Adding Confidence Intervals (CIs) | 1.4 Customizing pallete, title, and modifying default realized effect size calculations | 1.5 Customizing individual graphs | 2. GLM object | 2.1 Data and fitted object | 2.2 Default side-by-side plots, CIs and 99% confidence interval | 3 GLMNET model objects | 3.1. Data | 3.2 LASSO - data preparation and model | 3.3 Checking correlations for numeric variables of interest | 3.4 Default LASSO plots with custom realized effect size | 3.5 Modifying idividial plots and arraning them back together | 3.6 Post Selection Inference | 3.6.1 Data | 3.6.2 Data observations | 3.6.3 A note about rounding | 3.6.4 Model | 3.6.5 A note on data scaling and centering in relation to glmnet objects | 3.6.6 LASSO regression with CIs and custom realized effect size | 3.6.7 A note on Selective Inference | 3.6.7 GLMNET with penalty factor and CIs | 3.6.8 A second note on data scaling and centering in relation to fixedLassoInf objects | 3.6.9 Post selection inference with CIs and penalty factors

LIBRARIES | EXAMPLES | 1. Data | 2. Function | 3. Visualization | 3.1 Default Plots | 3.2 Customization | 3.3 Multiple Plots