eSMC

eSMC infers historical population admixture events from single-individual whole-genome data by extending the Pairwise Sequentially Markovian Coalescent (PSMC) framework to detect admixture signals.

Key Features:

• Admixture inference: Extends the Pairwise Sequentially Markovian Coalescent (PSMC) model to infer historical admixture events from single-individual genomes.
• Single-source data utilization: Infers admixture using genomic data from a single individual, enabling analyses when ancient or multiple-source datasets are unavailable.
• Simulation and validation: Validated with in silico simulations and real-world genomes using admixture times ranging from 5 kya to 100 kya and admixture ratios 1:1, 2:1, 3:1, and 4:1, with root mean square error reported across experiments.
• Application in diverse species: Applied to human individuals (Han and Tibetan) and domesticated species (donkeys and goats), yielding estimated admixture times of 60–80 kya for Han/Tibetan (25 years/generation), 40–60 kya for donkeys (8 years/generation), and 40–100 kya for goats (6 years/generation).
• Concordance with historical events: Inferred admixture times are consistent with known domestication and demographic events reported in the study.

Scientific Applications:

• Demographic history inference: Reconstructs historical admixture and effective population size dynamics from single genomes for demographic studies.
• Evolutionary biology and anthropology: Provides temporal estimates of admixture useful for studying human evolution and population interactions, as illustrated by Han and Tibetan analyses.
• Conservation genetics: Facilitates inference of admixture in domesticated and wild species (e.g., donkeys, goats) to inform conservation and breeding histories.
• Domestication and migration studies: Supports dating of domestication-related admixture events and migration patterns through estimated admixture times.

Methodology:

Extends PSMC by analyzing the distribution of the most recent common ancestors between diploid alleles, uses in silico simulations with specified admixture times and ratios, and evaluates performance using root mean square error on single-individual genomic data.