Post-Doctoral Research Fellow

Fred Hutchinson Cancer Center is an independent, nonprofit organization providing adult cancer treatment and groundbreaking research focused on cancer and infectious diseases. Based in Seattle, Fred Hutch is the only National Cancer Institute-designated cancer center in Washington.

With a track record of global leadership in bone marrow transplantation, HIV/AIDS prevention, immunotherapy and COVID-19 vaccines, Fred Hutch has earned a reputation as one of the world's leading cancer, infectious disease and biomedical research centers. Fred Hutch operates eight clinical care sites that provide medical oncology, infusion, radiation, proton therapy and related services, and network affiliations with hospitals in five states. Together, our fully integrated research and clinical care teams seek to discover new cures to the world's deadliest diseases and make life beyond cancer a reality.

At Fred Hutch we value collaboration, compassion, determination, excellence, innovation, integrity and respect. Our mission is directly tied to the humanity, dignity and inherent value of each employee, patient, community member and supporter. Our commitment to learning across our differences and similarities make us stronger. We seek employees who bring different and innovative ways of seeing the world and solving problems.

The Bedford Lab has worked extensively in the field of viral evolutionary forecasting. In this context, we've developed models to estimate fitness of seasonal influenza variants from genetic sequence data and to then use fitness estimates to forecast variant frequencies (Huddleston et al., eLife, 2020). We've taken a similar approach to forecasting SARS-CoV-2 variants in applying multinomial logistic regression (MLR) to estimate variant fitness and to project frequencies forward in time (Abousamra et al., PLoS Comput Biol, 2024). This approach underlies our live SARS-CoV-2 evolutionary forecasts at nextstrain.org/sars-cov-2/forecasts. Our influenza forecasts are directly utilized by the World Health Organization in the twice yearly vaccine strain selection meetings for seasonal influenza.

Recent advances in deep learning, especially transformer-based language models for protein sequences (see ESM3) and DNA sequences (see Evo2), present exciting new avenues to enhance evolutionary predictions. These models, trained to predict residues or nucleotides based on sequence context, have potential to significantly improve predictions of variant fitness and evolution.

We have an opening for a Post-Doctoral Research Fellow in the Bedford Lab at the Fred Hutch Cancer Center to work on developing and applying DNA and protein language models to understand and forecast viral evolution.

In this role, you'll initially focus on incorporating state-of-the-art language models to assess and predict the fitness of influenza and SARS-CoV-2 variants, comparing these predictions to our established statistical models.
• A key aim is to leverage these advanced models to provide deeper insights than traditional "mutational load" metrics, which simply count the number of amino acid changes.
• Additionally, you will explore how embedding spaces derived from these language models could offer new perspectives on evolutionary processes (see Hie et al for an example of looking at semantic change via embedding).
• Beyond applying existing language model frameworks, you'll have opportunities to design novel model architectures to describe the process of sequence evolution.

MINIMUM QUALIFICATIONS:

• A PhD in biology, computer science or a related field when starting the position is required. A quantitative background is essential, though PhDs from diverse fields including biology, mathematics, statistics, physics and computer science are welcome.
• The ideal candidate will have experience working with deep learning models via PyTorch or other frameworks. However, candidates with more traditional experience in sequence data and phylogenetic approaches who are excited to dive into deep learning models are also strongly encouraged to apply.
• Candidates should have experience in at least one programming language
• a proven track-record of peer reviewed publications.

• The candidate must be responsible, organized, and able to independently pursue research projects.

To apply please also submit

1. cover letter that includes the names and contacts for three references and a short statement of research interests
2. a current CV
3. code samples or links to code on GitHub

The annual base salary range for this position is from $77,976 to $95,014, and pay offered will be based on experience and qualifications.

Fred Hutchinson Cancer Center offers employees a comprehensive benefits package designed to enhance health, well-being, and financial security. Benefits include medical/vision, dental, flexible spending accounts, life, disability, retirement, family life support, employee assistance program, onsite health clinic, tuition reimbursement, paid vacation (22 days per year), paid sick leave (up to 30 calendar days per occurrence of a qualifying reason), paid holidays (up to 13 days per year), and paid parental leave (up to 4 weeks).