DENCode Transmission Probability Dataset and Ablation Results

DENCode is a probabilistic dengue virus (DENV) transmission inference framework that integrates haplotype-level viral genetics, clinical metadata, environmental covariates, and mosquito-biological constraints to estimate case-to-case transmission probabilities Pij. The dataset deposited here provides the full output of the DENCode transmission model, including epidemiological kernels, genetic similarity measures, and final weighted probability matrices. This release contains: 1. Transmission Probability Matrices For every case pair (i, j), the dataset includes: The epidemiological kernel Kij, incorporating temperature-dependent EIP, mosquito survival, spatial decay, and infectiousness profiles. The genetic similarity term Gij, computed using amino-acid haplotype patristic distances, mutation-hotspot penalties, and haplotype-abundance weighting. The final transmission probability Pij, proportional to Kij × Gij and normalized across all possible infectors for each case. 2. Haplotype-Derived Genetic Features Full haplotype-pair patristic distances for each case pair. Mutation annotations distinguishing hotspot vs. non-hotspot positions, using serotype-specific hotspot files. Haplotype abundance weights applied during computation of Gij. 3. Metadata Inputs Used by the Model Case-level clinical metadata (symptom onset, age, serotype). Geospatial coordinates for each case. Daily temperature values queried from external APIs (Open-Meteo / NASA POWER). Case-specific haplotype FASTA files. 4. Environmental and Mosquito-Model Components Temperature-derived extrinsic incubation period (EIP) functions. Mosquito mortality functions. Day-indexed kernel components used to construct Kij. 5. Reproducible Code and Pipeline Structure The full modeling pipeline, including scripts for: Parsing and encoding haplotype FASTA files Computing patristic distances Identifying hotspot mutations Calculating epidemiological kernels Combining components into weighted transmission probabilities 6. Ablation and Sensitivity Analysis The dataset includes ablation results evaluating the contribution of individual model components (Kij, Gij, hotspot weighting, temperature modulation, and mosquito-biological parameters) to the final transmission probability estimates. These files are provided for reproducibility and methodological transparency. is available at the associated GitHub repository:https://github.com/DENVGenomics/DENcode Purpose of the Dataset This dataset supports research on DENV transmission dynamics by providing a reproducible, haplotype-resolved set of probabilities that can be used for epidemiological inference, transmission network reconstruction, phylodynamics, or as a benchmark for future genetic-epidemiological modeling frameworks.