Decoding the fibromelanosis locus complex chromosomal rearrangement of black-bone chicken: genetic differentiation, selective sweeps and protein-coding changes in Kadaknath chicken

Black-bone chicken (BBC) meat is popular for its distinctive taste and texture. A complex chromosomal rearrangement at the fibromelanosis (Fm) locus on the 20th chromosome results in increased endothelin-3 (EDN3) gene expression and is responsible for melanin hyperpigmentation in BBC. We use public long-read sequencing data of the Silkie breed to resolve high-confidence haplotypes at the Fm locus spanning both Dup1 and Dup2 regions and establish that the Fm_2 scenario is correct of the three possible scenarios of the complex chromosomal rearrangement. The relationship between Chinese and Korean BBC breeds with Kadaknath native to India is underexplored. Our data from whole-genome re-sequencing establish that all BBC breeds, including Kadaknath, share the complex chromosomal rearrangement junctions at the fibromelanosis (Fm) locus. We also identify two Fm locus proximal regions (∼70 Kb and ∼300 Kb) with signatures of selection unique to Kadaknath. These regions harbor several genes with protein-coding changes, with the bactericidal/permeability-increasing-protein-like gene having two Kadaknath-specific changes within protein domains. Our results indicate that protein-coding changes in the bactericidal/permeability-increasing-protein-like gene hitchhiked with the Fm locus in Kadaknath due to close physical linkage. Identifying this Fm locus proximal selective sweep sheds light on the genetic distinctiveness of Kadaknath compared to other BBC.

Supplementary Figure 10. Pearson correlation coefficient test on the F ST comparison from ANGSD and popgenWindows.py method. F ST from popgenWindows.py at Y-axis and F ST from ANGSD at X-axis has a significant positive correlation with a correlation coefficient= 0.975 shown in blue circles.
Supplementary Figure 11. Geographical locations of black-bone chicken breeds used in this study. The map was generated using rworldmap, map, and mapdata packages implemented in R. Different shapes represent different breeds. All breeds in oval shape circles belong to China, and the zoomed view of the China map represents the sample location of each breed.    The long-read Nanopore data(SRR17968711, SRR17968712) of Silkie black bone chicken has been mapped to Gallus_gallus.GRCg6a genome assembly. Both haplotypes were differentiated based on the read group labeling using biostar214299, a program implemented in the jvarkit tool. Horizontal reads in light red represent haplotype-1 while reads in light green represent Haplotype2.
Supplementary Figure 26a. Screenshot of UCSC genome browser of two haplotype alleles at 24 positions of Dup1 region, which were identified using Nanopore long reads. Haplotype-1 of Dup1 spans from Flank1 to inverted Dup2 (i.e., Flank-1 + Dup1 + (inverted Dup2)). The purple lines represent overlapping reads of Nanopore containing Dup1 haplotype-1 alleles. The same purple lines represents the read from the end of haplotype-1 of Dup1 to the end of haplotype-2 of Dup2. Haplotype-2 spans Dup1 from haplotype-2 of Dup2 to the Int region (i.e., (inverted Dup2) + Dup1 + Int). The blue lines represent the overlapping reads of Nanopore containing haplotype-2 alleles. The same blue line represents the same read from the start of haplotype-2 of Dup1 to the start of haplotype-2 of Dup2. Both haplotypes of Dup1 are connected to the single haplotype of Dup2 (i.e., haplotype-2 of Dup2) but at different ends. The long-read Nanopore data(SRR17968711, SRR17968712) of Silkie black bone chicken has been mapped to Gallus_gallus.GRCg6a genome assembly. Both haplotypes were differentiated based on the read group labeling using biostar214299, a program implemented in the jvarkit tool. Horizontal reads in light blue represent haplotype-1 while reads in light red represent Haplotype2.

Supplementary
Supplementary Figure 28. Screenshot of UCSC genome browser of two haplotype alleles at 25 positions of Dup2 region, which were identified using Nanopore long reads. Haplotype-1 of Dup2 spans from Int to Flank2 (i.e., Int + Dup2 + Flank2). The grey lines represent overlapping reads of Nanopore containing Dup2 haplotype-1 alleles. Haplotype-2 of Dup2 spans from the start of haplotype-2 of Dup1 to the end of haplotype-1 of Dup1 region (i.e., Dup1 +(inverted Dup2)+ Dup1). The red lines represent the overlapping reads of Nanopore containing Dup2 haplotype-2 alleles. The same red line represents the span of the same reads from the start of haplotype-2 of Dup1 to the start of haplotype-2 of Dup2. Similarly, the same reads span the end of haplotype-1 of Dup1 to the end of haplotype-2 of Dup2.        Supplementary Figure 47a. Phylogenetic relationship of Dup1 and Dup2 region for 101 individuals using a variant call from bcftools has been shown. Using vk phylo with nj method used for phylogenetic construction. Red color nodes represent the black-bone chickens, and black color nodes represent the non-black-bone chickens.

Supplementary
Supplementary Figure 47b. Phylogenetic relationship of Dup1 and Dup2 region for 101 individuals using a variant call from bcftools has been shown. Using vk phylo with UGPMA method used for phylogenetic construction. Red color nodes represent the black-bone chickens, and black color nodes represent the non-black-bone chickens.    Figure 51. The origin of black-bone phenotype (Fm locus rearrangement) and distribution of BBC breeds: The common origin of black-bone chicken through a single event at Fm locus rearrangement has been shown. The grey color arrows represent the distribution of different BBC breeds from a common ancestor. The purple color arrow represents the origin of Yeonsan Ogye from Chinese black-bone breeds. The two-sided blue color arrow represents the gene flow between different BBC breeds and various native and commercial breeds. 4 11 5 Supplementary Figure 61. Population-specific private allele count shown on Y-axis and X-axis represents the populations. The numbers of individuals in each population is mentioned at the top of each bar. The counts of population-specific private alleles are obtained after removing the common alleles from dbSNP.