Task categories and types for AI agents for life sciences

ℹ️
Date 2024-10-14 to -15
Location BioLabs Heidelberg and Online

Category 1: AI agents for computational modelling and simulation

• Special software requirements: https://github.com/copasi/basico Documentation
• Task type 1 and Task type 3 Data for computational modelling: ODE mathematical models in SBML format from BioModels e.g., insulin dynamics, IBD disease dynamics
• Task type 2 Computational IBD model for parameter fitting: paper and xml file
• Task type 3 Computational IBD model in non-standard format: paper and MATLAB code
• Task type 3 Ground truth for MATLAB to SBML conversion: paper, MATLAB equations and SBML file.
• Data for model annotation: cell ontology, ChEBI, and UniProtKB

🧑‍🏫
Lilija Wehling

Task type 1

• Description: Forward simulation of a mathematical model and reporting of the biomarker trajectories and predicted clinical efficacy
• Input: simulation parameters such as initial concentrations
• Output: time-course of simulation species

Task type 2

• Description: Reverse fitting of a mathematical model and reporting of the parameter ranges
• Input: time-course of species
• Output: fitted model parameters

Task type 3

• Description: Creating a mathematical model from scratch
• Input: Original article describing the mathematical model and list of equations
• Output: SBML model with annotated species

Category 2: AI agents for omics and foundation models

• Special software requirements: Cell2Sentence
• Data for analysis: cell by gene
• Other tools/analyses: differential gene set enrichment analysis using GO, UMAP

🧑‍🏫
Gurdeep Singh

Task type 1

• Description: Integration of multiple scRNA seq datasets, correction for batch effects, and annotation of cells
• Input: multiple cell x gene datasets for a particular disease (e.g., Rheumatoid Arthritis, Atopic Dermatitis, Inflammatory Bowel Disease, etx.)
• Output: UMAP visualization with cell annotation

Task type 2

• Description: Simulation of gene perturbation and reporting of the predicted differentially expressed genes using pathway enrichment analysis
• Input: cell x gene dataset for a particular disease; knockout gene list
• Output: list of differentially expressed genes and pathway enrichment analysis visualization

Category 3: AI agent for Biomedical knowledge graph reasoning and construction

• Special software requirements: PyTorch Geometric and available models through PyG, LLMGraphTransformer, and schema-agnostic graph foundation model (e.g., ULTRA)
• Biomedical knowledge graph dataset: PrimeKG specifically the subset used in STARK for textual Q&A
• Other data: PubMed for original articles
• Graph database: NetworkX

🧑‍🏫
Ahmad Wisnu Mulyadi

Task type 1

• Description: Knowledge graph Q&A and retrieval of the K-hop subgraph explanations
• Input: Natural language question (see subset used in https://arxiv.org/abs/2404.13207 for PrimeKG)
• Output: Ranked nodes answers and visualization of k-hop subgraphs

Task type 2

• Description: Disease knowledge graph construction from text using a text-to-graph model to construct the initial knowledge graph and a link prediction model to fill in gaps in the reconstructed knowledge graph
• Input: List of disease MeSH terms and associated articles from PubMed and list of nodes and edges (same as in PrimeKG)
• Output: NetworkX representation of the knowledge graph and visualization

Task type 3

• Description: Same as type 1 but including protein embeddings from https://www.uniprot.org/help/embeddings and additional vector similarity search of drug targets embeddings
• Input: Natural language question (see subset used in https://arxiv.org/abs/2404.13207 for PrimeKG)
• Output: Ranked nodes answers and visualization of k-hop subgraphs