Intern, Molecular Modeling & Informatics at Alkermes in Greater Boston

Summary: Alkermes is a global biopharmaceutical company focused on developing medicines for complex psychiatric and neurological disorders. The role involves supporting research efforts by building and evaluating machine learning models using chemical assay data to enhance drug discovery processes. Responsibilities: - Build and evaluate machine learning models using chemical structure and assay data, with an emphasis on regression and classification tasks relevant to drug discovery - Apply modern cheminformatics and ML toolkits including ChemProp, scikit-learn, and RDKit to generate molecular representations and predictive models - Characterize model performance using appropriate validation strategies, including assessment of experimental error, noise, and upper bounds on achievable performance - Analyze internal assay datasets to understand data quality, variance, and implications for model reliability - Use model outputs to help inform decisions around potentially relevant chemical space and compound prioritization - Communicate results clearly through written summaries, figures, and presentations to technical stakeholders - If suitable this position will also contribute to the development of reusable, well documented modeling components that can be integrated into existing internal platforms and workflows Required Qualifications: - Enrollment in, or recent completion of, a degree program in computer science, data science, statistics, computational chemistry, cheminformatics, chemical engineering, or a related quantitative discipline Preferred Qualifications: - Fundamental understanding of machine learning concepts, including model training, validation, and evaluation - Experience working with Python for data analysis or modeling - Familiarity with structured scientific datasets and basic data cleaning / preprocessing - Ability to reason about uncertainty, error, and limitations in data‑driven models - Strong analytical thinking skills and attention to detail - Ability to communicate technical ideas clearly in written and verbal form - Hands‑on experience building machine learning models using ChemProp, scikit‑learn, or similar frameworks - Experience using RDKit or related cheminformatics libraries for molecular featurization and analysis - Strong Python coding skills, including experience writing modular, testable, and maintainable code - Experience developing reproducible modeling pipelines, including version control, parameter tracking, and evaluation workflows - Familiarity with experimental assay data, experimental variability, or model error analysis in scientific or engineering contexts - Exposure to software engineering best practices such as unit testing, code reviews, and documentation - Interest in applying machine learning to real‑world scientific decision making rather than purely theoretical modeling Required Skills: Machine Learning, Python, Chemoinformatics Important Skills: Data Analysis, Model Evaluation, Reproducible Modeling Nice-to-Have Skills: Analytical Thinking, Data Cleaning, Software Engineering Practices, Communication Skills