Skip to main content

Make your models shine!

References​

Workshop Series: Empowering Modelers with Jinkō​

Program structure​

  • Each workshop will last 2 hours, consisting of 30 to 45 minutes of instruction and demo and the remainder as hands-on practice.
  • Workshops can be taken in any order (beware though of non mandatory prerequisites) allowing flexibility for participants to tailor their learning path.
  • Materials provided: Workshop-specific assets, datasets, and access to a demo Jinkō environment with a Jinkō account.

This structure provides a comprehensive training program, addressing both technical and collaborative aspects of Jinkō’s platform for modelers.

1. Workshop: Literature Review & AI-Assisted Knowledge Extraction in Jinkō​

Objective: Help participants efficiently extract, organize, and utilize knowledge from the literature using Jinkō’s AI tools.

Overview:

  • Introduction to the Jinkō knowledge system for capturing scientific data​.
  • Using the AI tools for extraction
  • Best practices for systematic and transparent knowledge and data extraction with traceability​.

Key Topics:

  • Managing and organizing references.
  • Using AI to extract equation and data
  • Highlight key information and insert into the knowledge repository.
  • Creating a collaborative knowledge database for model building​.

Interactive Component:

  • Participants will perform a mock literature extraction using a predefined set of publications relevant to their projects.

2. Workshop: Model Import, Edition, and Creation from Scratch or with AI in Jinkō​

Objective: Equip participants with the skills to upload, edit, and create models in Jinkō, using its intuitive model-building tools.

Overview:

  • Uploading models (e.g., SBML or Simbiology format) into Jinkō​.
  • Editing and documenting existing models or building new ones from scratch​.
  • Visualizing and managing model components through dynamic representations.
  • Creating a model from paper using AI

Key Topics:

  • Autocompletion of parameters, annotations, and unit checks​.
  • Step-by-step guide on using tags for faster model use downstream.

Interactive Component:

  • Hands-on session where participants import and edit a basic pharmacokinetics (PK) model.

3. Workshop: Data Import & Model Calibration in Jinkō​

Objective: Teach participants how to bring external data into Jinkō, and calibrate models to ensure they accurately reflect real-world biology and clinical outcomes.

Overview:

  • Importing structured (clinical trial or lab) data
  • Formalized and unstructured data (literature, qualitative insights)​​ into scorings
  • Parametrizing models with genetic optimization algorithms (CMAES) and fitting virtual populations​​.

Key Topics:

  • Best practices for data handling and model calibration​.

Interactive Component:

  • Participants will import a dataset, calibrate a simple model, simulate and sub-sample a virtual population.

Recommended prerequisite: Workshop 1 and 2

4. Workshop: Running Clinical Trials in Jinkō​

Objective: Introduce participants to the process of setting up, simulating, and analyzing virtual clinical trials using Jinkō.

Overview:

  • Introduction to trial simulations: setting up protocol arms, adjusting doses, and adding variability​​ via a virtual population.
  • Running scalable simulations and tracking progress in real-time.
  • Population calibration via sub-sampling of completed trial results

Key Topics:

  • Trial design and optimization​​.
  • Result visualization: Kaplan-Meier survival curves, hazard rates, and endpoint analysis​.
  • Creating and fine-tuning virtual populations for clinical trial simulations​.

Interactive Component:

  • Simulation of a simplified phase III trial and comparison of trial arms.

Recommended prerequisite: Workshop 2

5. Workshop: Collaborating with Non-Modeling Experts & Onboarding Colleagues​

Objective: Facilitate interdisciplinary collaboration and streamline the onboarding process for teams not expert in modeling.

Overview:

  • How to present and document models for broad accessibility and safety​.
  • Collaborative features in Jinkō: shared model annotations, comments, and audits​.
  • Communicating results effectively to non-modeling stakeholders.
  • Prepare a simulation & analysis environment for non-modeling stakeholders.

Key Topics:

  • Presenting complex models for internal and external team presentations.
  • Collaborative workflows in Jinkō for interdisciplinary teams​.

Interactive Component:

  • Participants will create a shared project, assign roles, and collaboratively edit, run and analyze a trial.

Recommended prerequisite: Workshop 4

6. Workshop: Leveraging Jinkō’s API & Cookbooks for Programmatic Access​

Objective: Enable participants to use the Jinkō API for more flexible, automated workflows, integrating models with external tools.

Overview:

Key Topics:

  • Using the API to programmatically manage trial resources (disease model, population, protocols, ..), launch simulations and analyze results.
  • Integrating Jinkō with external environments like machine learning pipelines or visualization tools.

Interactive Component:

  • Set up a Jupyter-based workbench and an API key to interact with a project on Jinko
  • Learn by doing: create a population and a trial, launch a distributed simulation, analyze and visualize results.
  • Play with practical examples of real pipelines & integration.

Recommended prerequisite: Workshop 4