Multi-source Datasets and Modelling Code Supporting ‘Mapping Pesticide Mixtures to Cancer Risk at Country Scale with Spatial Exposomics’

Multi-source dataset, comprising:
(i) hard data file (.tif) underlying the high-resolution (100 m) nationwide environmental pesticide risk map across Peru (2014–2019), generated using a process-based model simulating the environmental behaviour of 31 pesticide active ingredients;
(ii) cancer registry data from the National Cancer Institute of Peru (INEN; 2007–2020), stratified by fifth-level Developmental Lineage Classification of Neoplasms, with ICD-10 codes and case counts;
(iii) district-level cancer standardised incidence ratios (SIRs > 1) across Peru, stratified by developmental lineage;
(iv) district-level relative cancer risks [RRs with lower bound of the 95% credible intervals (CI) > 1] across Peru, associated with modelled environmental pesticide exposure risk;
(v) sample enrichment scores (SES) computed from tumour and non-tumour liver tissues in cohorts from Peru (GSE111580, GSE136247), France (GSE62232), Taiwan (GSE45436), and Turkey (GSE17548), based on gene expression signatures indicative of exposure to non-genotoxic agents, hepatitis B virus, steatosis, alcohol, and foodborne genotoxins, as well as a gene set representing lineage-specific master transcription factors implicated in hepatobiliary carcinogenesis.

Modelling code, including:
(vi) R script for generating the process-based pesticide exposure risk map in Peru;
(vii) R script for integrated nested Laplace approximation (INLA) modelling of lineage-informed, pesticide-associated cancer risk;
(viii) Supplementary .zip archive containing additional files required to run the code and the test version.