Disentangling river and swamp buffalo genetic diversity: initial insights from the 1000 Buffalo Genomes Project

Abstract

More people in the world depend on water buffalo for their livelihoods than on any other domesticated animals, but its genetics is still not extensively explored. The 1000 Buffalo Genomes Project (1000BGP) provides genetic resources for global buffalo population study and tools to breed more sustainable and productive buffaloes. Here we report the most contiguous swamp buffalo genome assembly (PCC_UOA_SB_1v2) with substantial resolution of telomeric and centromeric repeats, ∼4-fold more contiguous than the existing reference river buffalo assembly and exceeding a recently published male swamp buffalo genome. This assembly was used along with the current reference to align 140 water buffalo short-read sequences and produce a public genetic resource with an average of ∼41 million single nucleotide polymorphisms per swamp and river buffalo genome. Comparison of the swamp and river buffalo sequences showed ∼1.5% genetic differences, and estimated divergence time occurred 3.1 million years ago (95% CI, 2.6-4.9). The open science model employed in the 1000BGP provides a key genomic resource and tools for a species with global economic relevance.

Keywords: SNP panel; buffalo genomics; carabao; structural variants; whole-genome sequencing.

Figure 1:

The female swamp buffalo from Kalinga Province, Philippines, was selected for whole-genome assembly.

Figure 2:

Comparison of gaps, major repeats, telomeric repeats, and centromeric repeats compared to other assemblies. (A) Barplot of the number of gaps per chromosome displaying the low number of gaps of the PCC_UOA_SB_1v2. (B) Violin plot for swamp and river buffalo genomes of repeat lengths >2 kb for LINE/L1, LINE/RTE-BovB, and satellite/centromeric repeats. The boxplot inside shows the quartile range and median. (C) Barplot of the centromeric satellite repeat families found in the tentative centromeric region of each chromosome. (D) Bedgraph for the telomeric signals of the 3 highly contiguous water buffalo assemblies. Telomeric count is equal to 1 unit of TTAGGG/CCCTGG. The red arrow represents the possible misassembly in chromosome 1 of UOA_WB_1.

Figure 3:

The phylogenetic tree of 8 species using single-copy ortholog genes indicating estimated time divergence and confidence interval from the present in Mya.

Figure 4:

Upset plot of the intersection of different types of SVs identified in water buffalo assemblies when aligned to PCC_UOA_SV_1v2 (swamp type), which shows the number of shared and unique SVs between different water buffalo assemblies.

Figure 5:

The first phase of the 1000BGP. (A) Bar graph of the average number of SNPs with standard deviation in swamp and river buffalo aligned with PCC_UOA_SB_1v2 (swamp type) and UOA_WB_1 (river type). (B) PCA plot using the swamp buffalo reference genome (PCC_UOA_SB_1v2) showing clear clustering of the swamp and river buffaloes. (C) Venn diagram of the number of autosomal SNPs in the 90K SNP buffalo genotyping array that are shared and specific for each water buffalo type. The large light peach circle shows the number of SNPs found in the 90K array; among these SNPs, the dark peach color shows river-specific SNPs, blue shows swamp-specific SNPs, and green shows SNPs shared for both types. (D) Histogram plot of the SNPs in the 90K SNP buffalo genotyping array, including both specific and shared SNPs per MAF value binned at 0.01.