Translational Science Intern

Join a Legacy of Innovation 125 Years and Counting!

Daiichi Sankyo Group is dedicated to the creation and supply of innovative pharmaceutical therapies to improve standards of care and address diversified, unmet medical needs of people globally by leveraging our world-class science and technology. With more than 125 years of scientific expertise and a presence in more than 20 countries, Daiichi Sankyo and its 18,000 employees around the world draw upon a rich legacy of innovation and a robust pipeline of promising new medicines to help people. In addition to a strong portfolio of medicines for cardiovascular diseases, under the Group’s 2025 Vision to become a “Global Pharma Innovator with Competitive Advantage in Oncology,” Daiichi Sankyo is primarily focused on providing novel therapies in oncology, as well as other research areas centered around rare diseases and immune disorders.   

We are currently seeking a Translational Science Intern from June 2026 – May 2027. This part-time position works for approximately 20 hours per week.

Responsibilities:

Project Title: Multimodal foundation models for integrating diverse biological data
The intern will use a multimodal AI framework to integrate and interpret diverse biological data, such as DNA sequencing, bulk RNA sequencing, single-cell multi-omics and imaging. The core objective is to use pretrained vast and diverse biological datasets to generate a holistic understanding of cellular states, gene regulation, and tissue architecture. We will demonstrate the utility of these models by applying them to a complex biomedical challenge, such as identifying novel biomarkers for cancer treatment. The project will address limitations of unimodal approaches, including noise sensitivity and incomplete biological context, by leveraging complementary and supplementary information across modalities. The models will incorporate interpretability methods to provide biological insights, foster new hypothesis generation, and accelerate biomarker discovery.
The focus is on technology evaluation and development to explore the latest methods out there and not limited to autoencoder, foundation model/transformer and graph neural network (GNN). In scope is the comparison with other conventional approaches for predictive biomarker discovery and patient subgrouping tasks.
Goals and Deliverables
The intern will be responsible for the following:
1. Literature Survey and Method Curation

• Conduct a structured review of recent ML/AI methods (published within the past 2-3 years) focused on biomarker discovery, subgroup and clinical efficacy prediction.

• Categorize methods by approach (e.g., deep learning, ensemble learning, survival analysis, multi-omics integration) and application area.

2. Prototype Implementation and Benchmarking

• Implement selected methods using internal or publicly available clinical datasets (e.g., TCGA, CPTAC).

• Compare performance using standard metrics (e.g., AUROC, concordance index, time-to-event analysis).

3. Pipeline Development

• Wrap up the development as a deployable package with easy-to-use configurable command line interfacing or notebook and provide user documentation.

• Ensure compatibility with internal data structures and analysis frameworks.

4. Reporting and Knowledge Transfer

• Deliver a final report and presentation summarizing findings, learnings, recommendations, and reproducible code.

Daiichi Sankyo, Inc. is an equal opportunity/affirmative action employer.  All qualified applicants will receive consideration for employment without regard to sex, gender identity, sexual orientation, race, color, religion, national origin, disability, protected veteran status, age, or any other characteristic protected by law.