Industry Consulting & Collaboration

Working with UCLA’s Institute for Commercialization & Industry Collaboration, our lab offers specialized skills to external parties through a fee-for-service model with greater flexibility as compared to traditional grant mechanisms.

Multivalent Binding Modeling for Optimizing Therapeutic Designs

Our lab offers advanced multivalent binding modeling services to help biotechnology companies optimize the design of multi-specific immunotherapeutic agents. Using a flexible multivalent binding framework (relevant publications), we can:

• Accurately model complex interactions between multi-specific agents and target cells
• Quantitatively characterize receptor engagement dynamics
• Predict therapeutic windows for various designs
• Analyze the impact of key design parameters on therapeutic activity and potential toxicity

Our services include:

• Data fitting and analysis of in vitro assay results
• Modeling of immune cell interactions in the presence and absence of tumor cells
• Comprehensive reports on parameter contributions to predicted outcomes
• Custom software for reproducing analyses

We work closely with clients to address specific scientific questions and provide actionable insights for therapeutic design optimization. We have worked with several organizations, from early startups to large pharmaceutical organizations, both as a grant or work for hire. Our expertise can help accelerate the drug development process and improve candidate selection.

Tensor Modeling and Big Data Exploration

Our lab specializes in applying innovative tensor-based approaches to analyze complex, multidimensional biological datasets that traditional matrix methods cannot adequately address. By leveraging tensor decomposition techniques, we help researchers extract meaningful patterns and relationships from high-dimensional biomedical data. These techniques offer several substantial benefits to biomedical data analysis:

• Preservation of data structure and relationships that would be lost in matrix-based approaches
• Increased statistical power through proper modeling of experimental design
• Discovery of higher-order interactions not detectable with standard methods
• Dimensionality reduction while maintaining interpretability
• Robust handling of sparse, noisy, or incomplete datasets common in biological studies

Our potential services include:

• Multidimensional Data Transformation: Converting conventional datasets into tensor structures that preserve the natural relationships between experimental dimensions (e.g., drug concentration, time points, cell types, and molecular features)
• Custom Tensor Decomposition Algorithms: Developing specialized algorithms tailored to the unique characteristics of biomedical data, including handling missing values, incorporating biological constraints, accounting for noise, and ensuring scalability
• Interpretable Feature Extraction: Identifying key factors and interactions across multiple dimensions that drive biological phenomena, making complex results accessible to domain experts
• Integration of Multi-Omic Data: Combining diverse data types (genomics, proteomics, metabolomics, etc.) within unified tensor frameworks to reveal cross-platform patterns

We work collaboratively with clients to implement these advanced analytical approaches, providing accessible software tools, comprehensive documentation, and expert consultation to ensure tensor methods enhance your research capabilities.

Relevant Publications