The team’s strategy rested on a simple but powerful premise: if centuries of selection drove teat number upward in Taihu pigs, the fingerprints of that selection should still be visible in their genomes. To find those fingerprints, the researchers analyzed whole-genome resequencing data from 213 pigs, using Asian wild boars as a common ancestral reference population. Comparing the Taihu breeds against this ancestral baseline allowed them to distinguish variants that arose recently and spread under selection from variation that has simply drifted over time. The comparison is critical because domestication itself reshapes genomes, and without a proper ancestral anchor, signals of recent selection can be confused with the deeper legacy of becoming a domestic animal in the first place.

Detecting selection is not a matter of a single statistical test, and the authors deployed three complementary approaches that each read the genome differently. Locus-specific branch length, or LSBL, measures how unusually differentiated a genomic region is in one population relative to others, flagging segments where allele frequencies have shifted dramatically. The integrated haplotype score, iHS, looks within a population for haplotypes that are unusually long for how common they are, a hallmark of an allele that rose to frequency so quickly that recombination has not yet broken it down. Cross-population extended haplotype homozygosity, XP-EHH, compares haplotype lengths between populations to catch selection that has already driven an allele nearly to fixation. Together, these methods cover the full sweep of selection scenarios, from variants still on their way up to those that have already won.

The genome-wide scan converged on 380 candidate genes sitting in regions shaped by selection in the Taihu breeds. When the researchers ran these genes through pathway enrichment analysis, a striking pattern emerged: they were significantly overrepresented in biological programs that govern mammary placode development, the embryonic process that lays down the primordia from which teats and mammary glands later form. Three signaling pathways dominated the list. The TGF-β pathway, the WNT pathway, and the PI3K-Akt pathway are all central orchestrators of embryonic patterning, controlling how cells communicate, proliferate, and organize into structures during early development. The candidate genes were also overrepresented in known quantitative trait loci, or QTLs, for teat number, meaning the regions flagged by selection overlapped with genomic intervals previously linked to the trait in independent mapping studies.

What elevates this study above a conventional selection scan is its second layer of evidence, drawn directly from the embryo itself. The researchers generated RNA-seq and ATAC-seq data from mammary placodes at embryonic day 26, a stage when these primordial structures are actively forming in the developing pig fetus. RNA-seq reveals which genes are being transcribed and at what levels, while ATAC-seq maps regions of open, accessible chromatin, the stretches of DNA where regulatory proteins can physically bind and switch genes on or off. By intersecting the selection-scan candidates with this developmental multi-omics data, the team could ask not just which genes show evolutionary signals, but which of them are actually active, and under regulatory control, in the very tissue whose development determines teat number.

This integrative filtering narrowed the field to 65 prioritized candidate genes. The researchers then pushed the evidence chain one step further with a phenome-wide association study, or pheWAS, which tests whether genetic variants in these genes show measurable associations with teat number in real animals. Thirty-seven of the 65 candidates cleared that bar, demonstrating significant links to the trait. The approach effectively triangulates from three independent directions: evolutionary history says a gene was under selection, developmental genomics says it operates in the relevant tissue at the relevant time, and association genetics says variation in it correlates with the trait itself. A gene that satisfies all three tests is a far stronger candidate than one that passes any single filter.

One gene stood out above the rest. BMPR1B, a bone morphogenetic protein receptor involved in transmitting signals from the BMP family of growth factors, was consistently supported by every line of evidence the team assembled. It carried selection signatures, showed differential expression in the embryonic mammary placode, fell within enriched developmental pathways, associated significantly with teat number in the pheWAS, aligned with teat number genome-wide association study results, displayed haplotype differentiation between high- and low-teat populations, and sat in regions of placode-accessible chromatin. BMP signaling is well known from developmental genetics as a regulator of epithelial organ formation, and the receptor’s convergence of evidence makes it the study’s strongest candidate for a gene that helps determine how many mammary placodes, and therefore teats, an embryo will form.

Beyond the coding sequence itself, the researchers hunted for regulatory variants, the non-coding DNA changes that alter when and where genes are expressed rather than changing the proteins they encode. Scanning the broader BMPR1B locus, they identified candidate regulatory variants located within chromatin regions that are accessible specifically in the embryonic mammary placode, suggesting these variants could act as developmental switches fine-tuning the receptor’s activity during placode formation. The authors are careful to note that the regulatory effects of these variants and their precise target gene remain to be determined experimentally. That caution matters, because regulatory variants are notoriously difficult to validate: a variant near a gene is not necessarily a variant that controls it, and confirming function will require laboratory work in developing embryos.

The study’s broader conclusion is that increased teat number in Taihu pigs is not the product of a single master gene but of a polygenic architecture, with many loci each contributing a modest effect that natural and artificial selection have cumulatively amplified. This finding fits a growing recognition in livestock genetics that production traits are typically built from many small genetic contributions rather than one dramatic mutation. It also explains why the trait has been so responsive to selection over generations of traditional breeding: with many genes each nudging the trait in the same direction, breeders selecting for higher teat counts had abundant raw material to work with, even without any knowledge of the underlying genetics.

The practical implications reach well beyond the Taihu breeds. Teat number constrains lactation capacity, and in modern pig production, where litters of fourteen or more piglets are now common, sows frequently cannot nurse their entire litter, forcing cross-fostering or supplementation. Candidate genes and regulatory variants identified here, BMPR1B foremost among them, offer molecular markers that breeders could incorporate into genomic selection programs to accelerate improvement of this reproductive trait in commercial lines. The study also demonstrates a methodological template: fusing population-scale genome scans with developmental multi-omics and association testing provides a way to move from statistical signals to biologically meaningful candidates, a workflow likely to be applied to other complex traits in livestock and, eventually, to questions of how domestication itself sculpts animal bodies.

Subject of Research: Integrative analysis of population selection signatures and embryonic mammary placode multi-omics reveals the genetic basis of increased teat number in Taihu indigenous pig breeds

Article Title: Integrative analysis of population selection signatures and embryonic mammary placode multi-omics reveals the genetic basis of increased teat number in Taihu indigenous pig breeds

Article References: Liu, C., Sun, C., Zhang, Y., Gao, N., & He, J. (2026). Integrative analysis of population selection signatures and embryonic mammary placode multi-omics reveals the genetic basis of increased teat number in Taihu indigenous pig breeds. BMC Genomics. https://doi.org/10.1186/s12864-026-13400-8

Image Credits: AI Generated

DOI: 10.1186/s12864-026-13400-8

Keywords: Integrative, analysis, population, selection, signatures, embryonic, mammary, placode, multi-omics, reveals, genetic, basis