Bioprocess Model Builder for Fast Customized Model Calibration
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing bioprocessing modeling requires significant modeling expertise, making it difficult for companies without in-house experts to adapt or create models for new biological systems or bioprocesses, necessitating external consultants each time.
Innovation Solution
A software-based Model Builder Module connected to an ontology database that assists users in defining model goals, selecting experimental data, composing models, and calibrating them using a toolset and algorithm, with features like automatic equation generation and expert-driven parameter suggestions, enabling non-experts to build customized models.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a customized scientific model is built for each new biological system or bioprocess, then the model accuracy and specificity are improved, but the time and resource consumption increase significantly
Solution Approach 1:
The patent applies preliminary action by pre-defining a library of standardized model templates and structures that cover common bioprocess types. These templates include pre-configured equations, parameters, and data requirements that can be directly applied or slightly modified for new processes, eliminating the need to build models from scratch each time.
Solution Approach 2:
The patent enables parameter changes by allowing users to systematically adjust model parameters to match specific biological systems while maintaining the overall model structure. The system provides guidance on which parameters need to be modified based on the specific application, reducing the complexity of model customization.
2Reliability
If detailed model engineering is performed to account for biological variability, then the model reliability is improved, but the complexity of the modeling process increases
Solution Approach 1:
The patent applies segmentation by dividing the model building process into distinct modular steps: data collection, model selection, parameter estimation, validation, and documentation. Each step is handled independently with specific guidance, making the overall complex process more manageable and less intimidating for users without extensive modeling expertise.
Solution Approach 2:
The patent introduces an intermediary computational tool that acts as a bridge between raw biological data and finalized models. This tool automatically performs intermediate tasks such as data preprocessing, equation formulation, and parameter optimization, reducing the burden on users while ensuring model reliability.
3Measurement precision
If external consultants are hired to build models, then the model quality is improved, but the cost increases
Solution Approach 1:
The patent enables self-service by providing an automated modeling system that guides users through the entire model building process without requiring external consultants. The system includes built-in expertise through automated parameter selection, model recommendation, and validation procedures that were previously only available through expert consultants.
Solution Approach 2:
The patent applies universality by creating a multi-functional modeling platform that can handle various bioprocess types (fermentation, cell culture, purification) using a common framework. This universal system eliminates the need to hire different specialists for different processes, reducing overall costs while maintaining model quality across applications.
4Measurement precision
If comprehensive data collection is performed for model calibration, then the model validity is improved, but the time and resource requirements increase
Solution Approach 1:
The patent applies partial action by identifying and collecting only the essential data elements needed for each specific model type, rather than comprehensively gathering all possible data. The system provides checklists of mandatory and optional data elements, allowing users to achieve sufficient model validity with minimal necessary data collection.
Solution Approach 2:
The patent uses preliminary action by pre-identifying data requirements for different model types before the actual modeling process begins. This allows users to collect only the necessary data in advance, avoiding unnecessary data collection efforts while ensuring all required information is available for model calibration.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Method for establishing and calibrating a customized software-based model for simulating a production process via a software-based Model Builder Module on a computer, wherein the Model Builder Module is connected to an ontology database that contains different domains, such as a bioprocessing data domain, a numeric simulation domain, and a mathematical model domains for the Model Builder Module, comprising the following steps of Defining the goal of the software based model, including information about the process the model describes, in particular process steps, cell line, molecule produced, variables the model predicts and threshold values of required model quality indicators; Selecting experimental data required for the calibration and the validation of the model in form of a list of experiments selectable in the Model Builder Module, wherein the context of the experiment data, which is stored in the ontology database matches the defined goal of the model; Composing the software-based model using the toolset proposed by the Model Builder Module which is specific to the respective process step the model describes; Calibrating the created software-based model with provided experimental data via the Model Builder Module; and Checking the validity and quality of the software-based model by performing a quality check of the Model using different mathematical analysis.