DataHowLab 26.5

This release transforms how you quantify process understanding, moving from simple robustness checks to a deep, statistical classification of your parameter space. 

• Sensitivity Analysis: Compute first, total, and second-order sensitivities simultaneously. This allows you to isolate the direct influence of inputs and reveal pairwise interactions that are often invisible in traditional one-factor-at-a-time approaches. 

• Feasibility Analysis: Move beyond “will it work” to “how likely is it to work.” The tool calculates the probability of simultaneously meeting all defined Objectives (CQAs/KPIs) across your parameter space, giving you a quantitative margin of safety. 

• Confidence Intervals: All sensitivity metrics are reported with confidence intervals, ensuring you know exactly when the model is reliable and when more data is required for a solid conclusion. 

• Objectives: A dedicated space to store information about the process relevant KPIs and CQAs.

• Recipe Templates: Store relevant operating conditions of your process.

• Parameter Ranges: Define the boundaries of exploration.

Pareto Insights is a major upgrade to our Design Optimization framework, shifting the power of choice from the algorithm back to the user. 

• Exploration Without Limits: Rather than receiving one “best” recipe, you now have access to all valid solutions on the Pareto Frontier, giving you the freedom to explore the entire design space based on your unique goals. 

• Data-Driven Guidance: Use the new Similarity Index to see how recipes compare to past experiments, making it easier to identify truly novel combinations or focus on familiar, high-performance areas. 

DataHowLab supports mammalian, microbial, gene therapy and mRNA process formats. We have extended our mRNA capabilities to include downstream processing, enabling end-to-end mRNA development. 

• IVT: Analyze and model mRNA in vitro transcription processes directly in the platform.
• PTC: Capture mRNA-specific process characteristics.
• Filtration: Analyze and model Tangential Flow Filtration (TFF) directly in the platform.
• Mechanistic Insights: Build propagation models for the Circulation, Ultrafiltration, and Diafiltration phases.
• Process Dynamics: Capture complex behaviors like cake formation and product concentration profiles to design more robust filtration stages. 

We’ve automated technical configurations so you can focus on scientific decisions rather than software settings. 

• Automated Model Training (No CV): Cross-validation settings are now handled automatically. The software applies the optimal approach based on your specific data, removing manual friction. 

• Actionable Diagnostics: New Out-of-Memory messaging detects failures caused by large datasets and provides concrete next steps for recovery, reducing guesswork. 

• Streamlined Connectors: With the latest improvements to Historical Connectors, the same import now completes in a matter of seconds, pulling data directly from their enterprise data lake into DataHowLab.
   

Efficiency updates to help you find information faster and maintain your workspace. 

• Advanced Table Filters: Quickly narrow down views to find relevant entries without excessive scrolling. 

• Type-to-Filter Dropdowns: Direct text input in dropdowns allows for faster navigation in large projects with many models or experiments. 

• Revised In-App Manual: A completely restructured layout and search functionality to provide answers right where you need them.
  
• User Deletion Update: A UX update and bug fix for the user deletion process to ensure smoother administrative management.