Bioinformatics Machine Learning Intern at RefinedScience in Remote

Summary: RefinedScience is dedicated to advancing care through innovative science and data integration. They are seeking a highly motivated Bioinformatics Machine Learning Intern to contribute to projects in single-cell biology and precision medicine by applying machine learning techniques to biological datasets. Responsibilities: - Analyze single-cell and multiomics datasets to extract biological insights supporting precision medicine and drug development programs - Apply and evaluate machine learning and deep learning approaches to single-cell data for tasks such as cell type classification, biomarker discovery, and patient stratification - Explore and prototype generative AI and LLM-based approaches to accelerate biological data interpretation and scientific workflows - Collaborate with scientists, clinicians, and data scientists to design and execute data-driven research projects - Document and optimize computational workflows following reproducible research best practices - Present findings through technical reports, visualizations, and presentations to cross-functional teams Required Qualifications: - Current Ph.D. candidate in Bioinformatics, Computational Biology, Computer Science, Biostatistics, or a related quantitative field - Single-cell omics experience: Demonstrated ability to process, analyze, and interpret single-cell data (scRNA-seq, scATAC-seq, CITE-seq, or spatial transcriptomics) using frameworks such as Scanpy/scverse, Seurat, or Bioconductor - Machine learning expertise: Applied experience developing and evaluating ML/deep learning models on biological data, including neural network architectures (GNNs, transformers, autoencoders), model selection and benchmarking, and integration of ML approaches into analytical workflows - Programming proficiency: Python and/or R for data analysis, statistical modeling, and visualization - Statistical foundation: Understanding of statistical methods for biological data (hypothesis testing, differential expression, multiple testing correction, clustering) - Strong problem-solving skills and ability to communicate complex insights effectively Preferred Qualifications: - Experience with deep learning frameworks (PyTorch, TensorFlow, JAX) - Familiarity with graph neural networks, attention mechanisms, or transformer architectures applied to biological data - Experience with ML experiment tracking and reproducibility (MLflow, Weights & Biases) - Exposure to representation learning, variational autoencoders, or contrastive learning methods - Familiarity with scikit-learn, XGBoost, or similar ML libraries - Interest in or experience with LLMs, RAG systems, or agentic AI tooling - Experience with multimodal single-cell integration (Seurat WNN, scvi-tools/MultiVI/totalVI, Muon) - Familiarity with spatial transcriptomics analysis (Squidpy, cell2location, nf-core/spatialvi) - Experience with cell-cell communication inference (CellChat, NicheNet, LIANA) - Knowledge of drug-gene interaction resources (CMap/LINCS, OpenTargets, ChEMBL) - Familiarity with Linux/Unix CLI and version control (Git/GitHub) - Experience with containerization (Docker, Singularity) and environment management (conda, venv) - Exposure to cloud computing platforms (GCP preferred) - Familiarity with workflow managers (Nextflow, Snakemake) - Adherence to best-practices for conduct reproducible computational research Required Skills: Single-cell omics, Machine learning expertise, Programming proficiency Important Skills: Statistical foundation, Deep learning frameworks, Bioinformatics experience Nice-to-Have Skills: Linux/Unix CLI, Version control (Git), Containerization (Docker), Cloud computing