Human-Mediated Dispersal Induces a Viable Reptile Hybrid Population: Caught in the Act

This repository contains all scripts, workflows, and intermediate files required to reproduce the genomic analyses of a recently discovered hybrid population of Cryptoblepharus skinks on Tromelin Island, as presented in "Human Mediated Dispersal Induces a Viable Reptile Hybrid Population: Caught in the Act". The study investigates the origin of this population and shows that it resulted from recent hybridization between Cryptoblepharus bitaeniatus and Cryptoblepharus caudatus, likely introduced in small numbers during logistical vessel movements. Genomic analyses indicate near equal ancestry proportions, strong founder effects, and divergence from both parental species, consistent with the establishment of a self sustaining hybrid lineage in an isolated insular system. The workflow is fully reproducible and processes genomic data from raw sequencing reads to population genomic and evolutionary inference. It is implemented using Nextflow pipelines together with bash, R, and Python scripts, and organized into modular analysis steps covering read processing, mapping, variant calling, and consensus generation. Downstream analyses include population structure inference using linkage disequilibrium filtering and NGSadmix, introgression and phylogenetic discordance testing using Dsuite, and admixture graph based f statistics. Hybrid classification is performed using a custom hybrid triangle framework combining ancestry proportions and interclass heterozygosity, complemented by RelateAdmix based identity by descent estimates and PCA using pcangsd. Phylogenetic and haplotype based analyses include mitochondrial haplotype network construction and genome wide FST estimation. Hybridization hypotheses are further tested using HyDe on linkage pruned datasets. Neutrality statistics are derived from site frequency spectra, and demographic history is inferred using fastsimcoal2 through model testing and bootstrap analyses. All analyses are organized into numbered modules with scripts, configuration files, and input lists. Conda environments are provided in the EnvsYML directory to ensure reproducibility. Modules follow a logical order from raw data processing to demographic inference, but can also be executed independently when dependencies are satisfied.