At Roivant, we are passionate about discovering and developing new drugs to impact patients' lives. Since its inception in 2014, Roivant has launched over 20 portfolio companies (Vants), overseen 5 successful IPOs, established a $3B partnership with a global pharma, built a pipeline of over 40 assets across various modalities and therapeutic areas, and delivered 8 successful phase 3 readouts.
Roivant is currently building new capabilities in drug discovery and expanding its existing development engine to become the world's leading tech-enabled pharmaceutical company. Roivant's drug discovery capabilities are driven by our computational discovery platform, which combines preeminent physics-based tools with deep expertise in machine learning to generate unprecedented predictive power that can tackle previously intractable discovery challenges. The tight integration of this computational platform with our experimental capabilities enables the rapid design and optimization of new drugs to address a wide range of targets for diseases with high unmet need.
We believe that the future of drug discovery lies in integrating predictive sciences, biology, and medicinal chemistry to accelerate the path to new medicines. This role is an opportunity to be an architect of this paradigm shift and generate transformative benefit for patients.
Roivant Discovery is looking for researchers and programmers in scientific computing with extensive experience in scientific programming, algorithm research, machine learning, or high-performance computing to join our platform team. Working closely with other platform team members, the candidate will develop and optimize advanced physics-based computational code—including but not limited to molecular dynamics simulations, free energy calculations, quantum chemistry, and machine learning models—to solve critical issues in drug discovery. Competitive pay, equity, strong perks, and a fun working environment, along with the opportunity to do cutting edge science to design better medicines, are all good reasons to join us!
- Develop—and implement in proprietary scientific software libraries—algorithms to enable new molecular physics and machine learning models. Active research areas include, but are not limited to:
- Collaborate with quantum chemists to develop machine learning models that predict accurate molecular energies at a fraction of the computational cost of quantum chemistry
- Collaborate with the molecular physics team to develop and implement new molecular interaction models within molecular simulation codes.
- Collaborate with the force field team to develop and train machine learning models to generate and improve force field parameters.
- Collaborate with the advanced simulations team to develop and implement new models to predict biophysical observables from molecular simulations and to guide the simulations with biophysical data
- Collaborate with the advanced simulations team to implement new enhanced sampling methods
- Develop and implement algorithms to accelerate our physics-based simulations and analysis
- Develop robust, scalable software that implement state-of-the-art algorithms in computational chemistry
- Optimize our scientific code for modern high performance and parallel computing architecture
- Highly motivated to develop computational methods and software for discovering better medicines
- B.S., M.S., or Ph.D. in computational physics/chemistry, physical chemistry/chemical physics, applied mathematics, computer science, or related fields
- Extensive programming experience in implementing and optimizing numerical methods (C/C++ and Python preferred)
- Excellent communication skills and strong team player
Additional Desirable Qualifications:
- Experience working with a diverse team on an ambitious project
- Experience in molecular dynamics simulations, Monte Carlo simulations, or other computational physics simulations
- Experience developing machine (such as random forests, support vector machines, deep neural networks) models, familiar with latest developments in machine learning methods.
- Extensive experience with common machine learning tools, such as Scikit.learn, Tensorflow, Pytorch, etc..
- Experience with high performance computing and parallel programming (e.g. MPI, openMP, CUDA)
- Experience in statistical modeling or bioinformatics