Director, Oncology Data Science

Merck Merck · Pharma · MA

Director of Oncology Data Science at Merck, focusing on multi-modal biomarker discovery using deep learning and AI methods on large clinico-genomic datasets. The role involves analyzing diverse tumor profiling data to inform oncology pipeline decisions and collaborating with AIML teams on foundation models.

What you'd actually do

  1. Enable reverse translation from large multi-modal clinical datasets to inform biomarker discovery and combination strategies in molecularly defined patient populations with high unmet medical need.
  2. Analyze, summarize and visualize the findings from large multi-modal clinico-genomic datasets which include bulk RNAseq, WES/WGS, imaging, epigenetic profiling, single-cell RNAseq, and proteomics data from oncology clinical trials and real-world datasets.
  3. Leverage advanced deep learning and AI methods and multivariate predictive modeling to discover novel biomarkers predictive of clinical outcomes.
  4. Effectively collaborate with AIML teams working on development of foundation models trained on large cohorts of human data.
  5. Present data analyses to inform and interact with stakeholders representing a wide span of internal organizations, including early discovery, translational and clinical development teams.

Skills

Required

  • Master's (with 10+ years) or Ph.D. (with 6+ years) in quantitative discipline such as Engineering, Applied Physics/Mathematics, Bioinformatics, Computational Biology or related field with a significant computational and statistical component and relevant experience in pharma, biotech or academic setting.
  • Experience in applying computational methods in cancer biology.
  • Demonstrated expertise in the application of methods of statistical learning and data mining to the integrative analysis of multimodal, high-dimensional tumor profiling datasets in the oncology and immuno-oncology context.
  • Hands-on analysis experience with the application of machine learning algorithms to large clinico-genomic, genetic and immunogenomic real-world and clinical trial datasets.
  • Extensive experience and demonstrated expertise to code in scientific computation environments with adoption of best practices for reproducible data analyses.
  • Strong communication and presentation skills; ability to guide and influence decisions through use of data-driven hypotheses; attention to detail.
  • Independent, flexible and collaborative mindset

Nice to have

  • Demonstrated experience with analysis of large-scale multi-modal data originating from clinical trials and real-world datasets.
  • Deep understanding of the major concepts of cancer biology as represented in multi-modal molecular and imaging data.
  • Experience within a matrixed industry environment and ability to effectively collaborate with colleagues from a wide range of disciplines.
  • Record of publishing in high profile scientific journals

What the JD emphasized

  • Experience in applying computational methods in cancer biology
  • Demonstrated expertise in the application of methods of statistical learning and data mining to the integrative analysis of multimodal, high-dimensional tumor profiling datasets in the oncology and immuno-oncology context.
  • Hands-on analysis experience with the application of machine learning algorithms to large clinico-genomic, genetic and immunogenomic real-world and clinical trial datasets.
  • Extensive experience and demonstrated expertise to code in scientific computation environments with adoption of best practices for reproducible data analyses.

Other signals

  • multi-modal biomarker discovery
  • deep learning and AI methods
  • multivariate predictive modeling
  • foundation models