Genomic variation responding to artificial selection on different lines of Pekin duck

Abstract

Understanding the genomic variation in Pekin duck under artificial selection is important for improving the utilization of duck genetic resources. Here, the genomic changes in Pekin duck were analyzed by using the genome resequencing data from 96 individual samples, including 2 conservation populations and 4 breeding populations with different breeding backgrounds. The population structure, runs of homozygosity (ROH), effective population number (Ne), and other genetic parameters were analyzed. The breeding populations showed lower genetic diversity compared to the conservation populations. Maple Leaf duck and Cherry Valley duck retained low genetic diversity compared to other breeding populations, with Cherry Valley duck showing the lowest diversity and the highest inbreeding coefficient. This suggested that Cherry Valley and Maple Leaf ducks have undergone intensive selection compared to other breeding populations. By the analysis of runs of homozygosity (ROHs), some genes (e.g., IGF1R) associated with growth traits were identified. By the analysis of the selection signal, strong selection characteristics in certain genomic regions during the breeding of Peking duck across different selected lines were observed. In addition, copy number variations (CNVs) in Pekin duck populations were analyzed. Six regions of interest were identified, containing RPA1, DOT1L, SLC25A42, RALYL, TRPA1, and IGFBP2. Furthermore, the allele frequency distribution of these genes showed significant differences between breeding populations and conservation populations, indicating that these candidate genes could have undergone strong selection pressure during long-term selection for improved production. These findings contribute to a deeper understanding of the distinct evolutionary processes in Pekin ducks under artificial selection and provide valuable insights for future breeding strategies.

Keywords: Genetic diversity; Intensive selection; Pekin duck; Weight.

Figure 1

The information of genetic variation and population structure in Pekin duck populations. a, Venn diagrams illustrate the shared SNPs in the six populations. b, Population structure analysis (K = 2). c, Principal component analysis revealing genetic differentiation of 7 populations using SNP data. The conservation populations are displayed as dots, Mallard is displayed as triangle symbols, PK3 and PK4 are denoted by plus sign, and ML and CV are marked as asterisks. d, Neighbor-joining tree.

Figure 2

LD decay, historical generations effective population size, genetic diversity and inbreeding coefficients (F_ROH) in Pekin duck populations. a, LD decay in different populations. b, Effective population sizes of different populations. c, Nucleotide diversity (pi) and genetic differentiation (F_ST) across the six populations. The value in each circle represents a measure of nucleotide diversity for this breed, and the value on each line indicates genetic differentiation between the two breeds. d, Inbreeding coefficients in populations.

Figure 3

Manhattan plot of the percentage occurrence of SNPs in ROHs for each population.

Figure 4

Overlapped genes of ROH islands identified in breeding populations compared to the conservation population. a, Venn diagrams illustrate the shared genes compared to the conservation population. b, Distribution of runs of homozygosity (ROH) on IGF1R per individual.

Figure 5

Genome-wide screening of selected copy number variations (CNVs). a, The relative frequency difference (RFD) between the conservation populations and the breeding populations is plotted according to the position of autosomes. b, Frequency changes in candidate CNVs contained on DOT1L in each population. c, the relative frequency difference (RFD) between PK3 and PK4 is plotted according to the position of autosomes. d, Frequency changes in candidate CNVs contained on TRPA1 in each population.

Figure 6

Selection signal analysis in the Pekin duck populations. a, Manhattan plot of selected regions between the breeding and conservation populations, followed by a comparison of PK3 and conservation populations, PK4 and conservation populations, ML and conservation populations, CV and conservation populations. Pairwise fixation index values (F_ST) are calculated in 10-kb sliding windows and 5-kb steps. b, Venn diagrams illustrate the shared candidate genes in the six populations. c, Venn diagrams illustrate overlapping genes screened by F_ST and XPCLR. d, Profile of mutant allele frequency changes for the six Pekin duck populations at the selection signal associated SNPs in ETF1 gene. e, The ETF1 gene showed a different genetic signature in breeding and conservation populations. The Nucleotide diversity (π) and Tajima's D values around the ETF1 locus. The ETF1 gene region is shown in gray.