Track IV EBP (Driven by Biodiversity Biobanks): Subtrack 1. Animals(J28020)

Foundation Models

1.Single-species deep model: Similar to Genos, trained on whole-genome datasets from a single model organism or an economically important animal species, enabling fine-grained characterization of species-specific genomic features.

2.Cross-species general model: Models trained on genomes from hundreds of species representing key evolutionary nodes, capable of learning deeply conserved evolutionary patterns and thereby substantially improving genome annotation of newly discovered species.

3.Multimodal genomic foundation model: Genomic foundation models extended to integrate and process additional data modalities beyond DNA sequences, such as transcriptomic, epigenomic, or phenotypic data.

4.Cross-species single-cell omics foundation model for animals: Large models trained on single-cell omics data across diverse animal species to capture conserved and species-specific cellular and regulatory architectures.

Vertical Applications

• 1.Novel genome assembly algorithms
• 2.Advanced genome polishing/correction algorithms
• 3.“One-click” automated genome annotation
• 4.Precision breeding and genetic improvement
• 5.Endangered species management
• 6.Mining of biodiversity genetic resources

Affiliated Branches

• BGI Research, Shenzhen
• BGI Research, Wuhan

Qualifications

Preferred academic background: Computer Science or Bioinformatics.

Programming skills: Proficiency in Python, with a solid understanding of basic algorithms and practical experience using Linux environments.

Data analysis and modeling: Demonstrated ability to perform statistical analysis and/or build machine learning models.

Literature proficiency: Ability to independently read and comprehend English-language scientific literature and technical documentation.

Research or project experience (preferred): Prior involvement in research or applied projects is highly desirable. Applicants are encouraged to provide evidence such as GitHub repositories or published papers.