Track II 10-Billion-Cell Alliance: Subtrack 4. Oncology(J28009)

Foundation Model Objectives

Develop a model capable of tumor identification, cellular state analysis, tumor evolution inference, prognosis prediction, and potential targeted therapy recommendation.

Key Components

• 1.Model architecture design
• 2.Multimodal unified semantic space
• 3.Learning task: Construction of virtual “digital tumor cells” to simulate tumor state transitions and therapeutic responses

Vertical Applications

• 1.Early tumor screening and risk monitoring
• 2.Minimal residual disease (MRD) monitoring and relapse warning
• 3.Companion diagnostic biomarker screening
• 4.Precision treatment decision support
• 5.Drug perturbation simulation and efficacy prediction

Affiliated Branches

• Southwest Branch (Chongqing)
• Beijing Branch (Beijing)
• Hangzhou Branch (Hangzhou)

(1) Preferred Background

Cancer Biology, Bioinformatics, Computational Biology, AI, Systems Medicine.

(2) Core Technical Skills

Multimodal deep learning; Programming & Data Analysis, Multimodal Data Analysis (scRNA-seq/scATAC-seq).

(3) Modeling Competency

Basic ability to develop computational models for tumor cell state characterization and subtype analysis. Interest in exploring digital tumor cell representations to model tumor state transitions and investigate potential treatment responses.

(4) Preferred Qualifications

Experience with cancer genomics, pathology AI, or multimodal biomedical data analysis; familiarity with large foundation models.

(5) Personal Traits

Strong curiosity for AI and biomedical research, willingness to learn across disciplines, and persistence in addressing complex scientific problems.