Animal genetics & breeding · comparative genomics

Jiaqi Wu 呉 佳齊 · ウー・ジャーチー

Assistant Professor, Hiroshima University
Graduate School of Integrated Sciences for Life & School of Applied Biological Science (Applied Animal & Plant Science) · Higashi-Hiroshima, Japan

I study how traits, breeding, and adaptation arise from large-scale genomic data across mammals, poultry and livestock, wild animals, and viruses — and I build methods that make large-scale comparative genomic analysis comparable and reproducible, so that the same branch really is being compared across genes.

01

About

I am an evolutionary genomicist in the Applied Animal and Plant Science program at Hiroshima University's School of Applied Biological Science, where I teach animal genetics and breeding. I work at the intersection of comparative genomics, molecular evolution, statistical genetics, and bioinformatics, extracting shared evolutionary principles from large-scale genomic data across mammals, birds and poultry, livestock, insects, and viruses.

To me, animal-breeding genomics and evolutionary genomics are one framework, not two. The same statistical-genomics toolkit that dates mammalian life-history evolution also speaks to poultry and livestock breeding — the origins, population structure, and diversity of native chickens and goats — and to the genomic monitoring of wild mammals for conservation and population management, as in my work on the Asian black bear, Japanese wolves, and bats. I connect the deep history of species divergence with within-species and within-breed diversity, adaptation, and selection on a common footing.

Trained in medicine (BS, Shandong University), bioinformatics (MS, Fudan University, with the late Prof. Yang Zhong and Prof. Masami Hasegawa), and statistical genetics (PhD, The University of Tokyo, with Prof. Hirohisa Kishino), I have since held positions at Tokyo Institute of Technology (JSPS Research Fellow), Tokai University School of Medicine (JST CREST), and Hiroshima University. A recurring theme in my methodological work is whether large-scale analyses are really comparing the same evolutionary object — and building the methods, benchmarks, and reproducible workflows that make sure they are.

02

Research

From animal genetics and breeding to method development — connecting the deep history of species divergence with within-breed diversity, adaptation, and selection on a common, auditable footing.

Phylogenomic methods

Branch comparability across genes

When taxa are missing or gene trees disagree, distinct species-tree branches can collapse or vanish. I develop split-based coordinate systems that make branch identity — and every kind of absence — explicit and reproducible.

Molecular evolution

Rates, life history & constraint

Molecular evolutionary rate as a window on mammalian life-history evolution, functional constraint, and intraspecific polymorphism — including a proposed post-K–Pg nocturnal bottleneck of placental mammals.

Viral genomics

Genome surveillance & variant dynamics

Design and analysis of viral genome-variation databases for the Genotype-to-Phenotype Japan (G2P-Japan) consortium — SARS-CoV-2 variant emergence and spread, plus HCV drug-resistance and KSHV transmission.

Animal genetics & breeding

Poultry, livestock & wild mammals

Population and comparative genomics for breeding and conservation — origins and structure of native chickens and goats, and genomic monitoring of wild mammals (Asian black bear, Japanese wolves, bats), extending to silkworm, noctuid pests, ratite birds, and beyond.

Current focus — SplitAligner

A branch-identity coordinate system for phylogenomics under missing taxa and gene-tree discordance. It defines branch identity through projected species-tree splits and decomposes branch absence into explicit, auditable categories, with a deterministic graph-oracle benchmark (Catnip10) and an R reference implementation in progress. Its branch coordinates are defined by canonical unrooted splits rather than display-root or positional node identities, so equivalent representations of the same weighted unrooted tree preserve the same biological branch identities.

Current focus — Catnip10

A deterministic 10-tip conformance benchmark for branch-wise workflows that file values on a fixed or shared reference-branch axis. An independent graph oracle supplies the truth object, and a workflow is scored on two axes: pruning-aware coordinate eligibility, and representation-stable coordinate identity and availability. SplitAligner builds coordinates; Catnip10 tests them.

03

Software

Open-source tools I develop and maintain. Methods, validation data, and reproducible workflows are released together.

SplitAligner

Perlphylogenomics2026

A branch-identity coordinate system that maps each gene tree onto species-tree splits under a locus's own taxon coverage, and names every kind of branch absence (structural, fusion, topology-induced). Ships the deterministic Catnip10 audit benchmark.

Catnip10

Python / RMITconformance benchmark

A deterministic ten-tip conformance benchmark for branch-wise workflows that file values on a fixed or shared reference-branch axis. An independent graph-contraction oracle supplies the truth object, and workflows are scored on two axes: pruning-aware coordinate eligibility, and representation-stable coordinate identity. Developed within the SplitAligner validation ecosystem.

SplitAlignerR

RGPL-3in development

The R reference-implementation track for SplitAligner. The current seed release ships the Catnip10 graph-oracle benchmark as audit-ready, documented package data with validation helpers; the empirical split-mapping engine is in active development.

SGV-Caller

PerlMIT2025

A SARS-CoV-2 genome-variation caller that builds local variation databases at nucleotide, codon, and amino-acid levels from GISAID data. Developed for the G2P-Japan consortium as an underlying data source for monitoring viral variants in Japan.

04

Publications

Selected peer-reviewed articles; full list on Google Scholar ↗. Add entries by editing the PUBS array in this file.

05

Contact

I welcome collaborations across comparative genomics, molecular evolution, statistical genetics, and bioinformatics — from method development and benchmarking to real-data analysis in animals, viruses, and beyond.