Machine Learning Scientist, Scientific Reasoning Models, AI for Drug Discovery

Roche, a leader in biotechnology, is actively seeking to enhance its drug discovery processes through the integration of advanced artificial intelligence (AI) and machine learning (ML) technologies. The company has established a new role for a Machine Learning Scientist within its AI for Drug Discovery (AIDD) group, specifically in the Foundation Models team at Genentech (gRED). This initiative is part of Roche's broader mission to innovate healthcare solutions and ensure that patients have access to transformative medicines.

Advancements in Drug Discovery

Roche's Research and Early Development organizations, encompassing both gRED and Pharma (pRED), are at the forefront of leveraging AI, data, and computational sciences to revolutionize the drug discovery and development landscape. By harnessing these technologies, Roche aims to accelerate research and development (R&D) processes, ultimately leading to improved health outcomes. A critical component of this strategy is the establishment of the Computational Sciences Center of Excellence (CoE), which serves as a unified group focused on utilizing data and AI to empower scientists in both pRED and gRED.

The AIDD group is particularly focused on applying cutting-edge machine learning techniques to drug discovery. The newly created position of Machine Learning Scientist will play a pivotal role in developing internal reasoning Large Language Models (LLMs) that are tailored to address specific challenges in drug discovery, such as biomolecular design. This role requires a unique blend of engineering and research skills to design and scale large machine learning systems effectively.

Key Responsibilities and Skills

The Machine Learning Scientist will be responsible for several key tasks that are essential to the success of Roche's drug discovery initiatives. These include designing and implementing large-scale distributed machine learning systems, writing robust and performance-critical code, and contributing to core infrastructure improvements. The scientist will also be tasked with developing strategies to enhance model performance on scientific tasks, which may involve complex reasoning and long-horizon task completion.

A significant aspect of the role involves translating biological and chemical domain knowledge into actionable machine learning objectives. This includes establishing training signals and evaluation criteria that align with the specific needs of biological research. The scientist will also collaborate closely with domain experts to design evaluation methodologies that assess model capabilities and curate high-quality data for training purposes.

Moreover, the Machine Learning Scientist will work in tandem with researchers to transition innovative ideas and prototypes into scalable, production-ready systems. This collaboration is crucial for ensuring that theoretical advancements in machine learning can be effectively applied to real-world drug discovery challenges.

Qualifications and Expectations

Candidates for the Machine Learning Scientist position are expected to have a strong educational background, with a BS/MS in Computer Science, Statistics, Mathematics, Physics, or a related quantitative field, along with at least two years of relevant work experience. Alternatively, a Ph.D. with zero to two years of experience is also acceptable. Key qualifications include expertise in developing and training large-scale machine learning models, a solid publication record in top-tier research venues, and strong software engineering skills.

Experience with molecular modalities, such as protein sequences and chemical graphs, is essential for the role. Additionally, applicants should have a public portfolio showcasing their research or contributions to open-source machine learning libraries, which underscores their expertise and commitment to the field.

The expected salary range for this position varies by location, with New York City salaries ranging from $141,100 to $262,100, and San Francisco salaries ranging from $147,600 to $274,000. Actual compensation will depend on various factors, including experience and qualifications. Roche also offers a discretionary annual bonus based on individual and company performance, along with comprehensive benefits.

Genentech, as an equal opportunity employer, emphasizes a commitment to diversity and inclusion in its hiring practices, ensuring that all candidates are treated based on merit and qualifications, free from discrimination.

Why it matters

The establishment of this Machine Learning Scientist position at Roche is significant for several reasons. Firstly, it highlights the growing importance of AI and machine learning in the pharmaceutical industry, particularly in drug discovery. By integrating advanced computational techniques, Roche aims to streamline R&D processes, which could lead to faster and more efficient drug development cycles.

For creators and technologists, this development underscores the potential for collaboration between AI and life sciences, paving the way for innovative solutions that can address complex biological challenges. The role's focus on translating domain knowledge into machine learning objectives exemplifies the need for interdisciplinary approaches in tackling real-world problems.

Furthermore, the emphasis on rigorous evaluation methodologies and collaboration with domain experts reflects a broader trend towards ensuring that AI applications in healthcare are not only innovative but also reliable and effective. As the field continues to evolve, the integration of machine learning in drug discovery could significantly enhance the capabilities of researchers and scientists, ultimately benefiting patients worldwide.