National Cancer Institute, Frederick, MD
About the position
The Center for Cancer Research (CCR) is home to nearly 250 basic and clinical research groups located on two campuses just outside of Washington, D.C. CCR is part of the National Cancer Institute (NCI) and makes up the largest component of the research effort at the National Institutes of Health (NIH). Centrally supported by long-term funding and a culture of complete intellectual freedom, CCR scientists are able to pursue the most important and challenging problems in cancer research. We collaborate with academic and commercial partners and advocacy groups across the world in efforts to prevent, diagnose and treat cancer and HIV/AIDS. The CCR research portfolio covers the full spectrum of biological and biomedical research. Our work ranges from basic to translational and clinical, and our clinical trials are conducted in the NIH Clinical Center, the world’s largest hospital dedicated to clinical research that offers a robust infrastructure to support CCR’s patients on an estimated 250 open studies. The success of CCR is grounded in an exceptionally strong discovery research program that provides the foundation for the seamless translation of insights from bench to bedside. Read more about CCR, the benefits of working at CCR and hear from our staff on their CCR experiences.
The Center for Structural Biology, the National Cancer Institute, National Institutes of Health, seeks a computational biophysicist to fill a permanent staff position at a Staff Scientist level. We are looking for an enthusiastic, talented and dedicated scientist to join our exciting research program. The detailed specific expertise and skills are listed in the following. Salary is commensurate with experiences and achievements.
- Proven record of deep understanding of Physics and physics instrumentation, physical phenomena, and iteration of the radiation/particles with matter, with background to understand the whole process of signal processing from the detector to the final data acquisition output.
- Deep understanding of complex detector resolution effects and a proven record of developing methods to simulate complex detector effects on simulated data, being able to describe real experimental data using simulations.
- Proven record of working with complex and extremely large datasets, in the order of petabytes of data and efficiently handling the data through High Performance Computing environment using bash, C/C++ and Python, writing efficient code.
- Experience with Object-oriented programming (OOP) (C++ and Python).
- Deep understanding of the process to analyze experimental data, background subtraction, detector efficiency corrections, statistical and systematical uncertainties evaluation, and simulation comparisons to drive physical insights.
- Proven record of dealing with datasets where the noise/background can be much larger than the actual signal.
- Proven record of developing novel methods for data analysis and simulations over different fields, such as SAXS, WAXS and AFM considering physical effects.
- Experience in analyzing data from Synchrotron Light Laboratories.
- Proven record of technical background applying AI to solve biological problems, such as:
- Training new Computer Vision models for object detection, classification, and instance segmentation, such as for cell classification/segmentation.
- Training and designing the architecture of Deep Learning models for RNA selection.
- Development and application of optimizer algorithms from scratch such as Gradient Descent and Instance Segmentation Analysis algorithms, with deep understanding of data augmentation, post-processing detection inference as well as non-maximum suppression (NMS).
- Working with AFM experimental images as well as with the simulation of AFM images and comparison to RNA models.
- Deep knowledge of hyperparameter tuning.
- Ablation studies.
- AUC studies.
- Learning curve studies.
- Proven record of using Machine Learning tools commonly used in the industry for machine learning model management such as MLFlow.
- Proven record of developing end-to-end Python packages in git repositories, using docker and containerization concepts, Continuous Integration, Unit Tests and the best software engineering practices for code documentation.
- Proven record of contribution to data engineering pipelines such as to archival/retrieval of data and metadata for data management purposes.
- Proven record of leading and mentoring Deep Learning projects.
- Proven record of developing workflow pipelines, or automatically converting generically written python/R scripts into workflows, such as in Common Workflow Language or Snakemake pipelines.
- Knowledge of how to build algorithms that can be parallelized on CPUs.
- Record of leading team efforts to deploy large software in HPC systems, making it available to other users and providing support.
- Some knowledge in system administration, managing group and user permissions, root access, server administration, installation and maintenance of conda environments and packages.
The NIH is dedicated to building a community in its training and employment programs and encourages the application and nomination of qualified women, minorities, and individuals with disabilities.