<aside> 💡

MObyDiCK (Mechanistic Target Optimization by Discovering and Controlling Network) platform is one of innovative and revolutionary products of our research from the last two decades, which incorporates systems biology and network control theories to identify first-in-class targets that can revert cancer and aged cells into normal and healthy cells [1-5]. Using this, various experimental data, including single-cell omics data, can be incorporated to construct context-specific mathematic models for conducting extensive computational simulations in order to decipher underlying mechanisms of cancer and aging.

</aside>

MObyDiCK Target Identification Service (MObyDiCK-TIS)

The MObyDiCK platform offers a cutting-edge Target Identification Service, designed to revolutionize the early stages of drug discovery. By leveraging advanced systems biology analysis and single-cell omics data, MObyDiCK constructs detailed gene regulatory networks that go beyond traditional correlation-based methods. This platform uncovers causal relationships between genes, allowing for the precise identification of first-in-class drug targets. Through comprehensive simulation analyses, MObyDiCK not only identifies potential targets but also provides deep insights into their mechanisms of action. MObyDiCK-TIS includes in vitro validation of targets by cell-based phenotype assay (e.g. proliferation, expression of marker gene, etc.). This unique approach accelerates drug target identification, reduces development costs, and increases the likelihood of success in bringing innovative therapies to market.

Features

<aside> 🔥

The MObyDiCK platform is based on a patented method (KR20230140402A) for pseudo-time analysis and network modeling. This patent outlines a novel method for constructing gene regulatory networks (GRNs) from single-cell omics data by utilizing pseudo-time analysis to infer causal relationships between genes, enabling precise identification of drug targets.

</aside>

• The platform utilizes single-cell omics data sourced from in vitro cell lines, patient-derived cancer organoid models, and publicly available cancer patient datasets. This type of data provides a detailed view of the individual states of cells, allowing the platform to extract temporal information by pseudo-time analysis without the need for longitudinal sampling [PMC1, PMC2].
• Pseudo-time analysis assumes that cells with similar gene activity have progressed through similar stages in a biological process. This temporal information helps the platform infer causal relationships by identifying which genes are activated earlier in the process and may be influencing the expression of downstream genes.
  • By tracking changes in gene expression along the pseudo-time trajectory, the platform can distinguish cause-and-effect relationships between gene interactions, rather than mere correlations.
• The platform leverages expertise in network modeling, using systems biology and AI to select key genes and infer their regulatory functions. The MObyDiCK platform employs Boolean modeling, where each gene is represented as either active (1) or inactive (0), simplifying the state of the gene within the GRN.
  • While Boolean modeling does not represent the full range of gene expression values, most genes exhibit bimodal distribution in their expression patterns, transitioning sharply between on and off states. This sigmoidal transition allows Boolean modeling to effectively capture the core dynamics of GRNs, providing useful insights despite the simplified binary framework.

<aside> 🔑

Causal relationships and regulatory functions are the core outputs of the MObyDiCK platform. These results provide a deeper understanding of disease progression, enabling the identification of driving factors and potential drug targets that are difficult to uncover using traditional methods such as differential gene expression or simple gene correlation analysis. By moving beyond correlation, MObyDiCK reveals the actual gene interactions and regulatory mechanisms, leading to more accurate and actionable discoveries in drug development.

</aside>

• These functions are integrated into a mathematical model of the GRN.
  • Due to the computational complexity of simulation analysis, each GRN model typically contains up to 30 nodes.
  • MObyDiCK can generate multiple GRN models (an ensemble model) from the same dataset, all of which reproduce the dataset with equal accuracy. Analyzing these ensemble models provides greater robustness to noise and compensates for the incompleteness of data, leading to more reliable results in identifying regulatory mechanisms.
  • The current GRN model is designed for a specific cell type or lineage, which means cell-to-cell communication is not included in the model. To address this, we plan to upgrade the MObyDiCK platform to extend the GRN model, enabling it to capture cell-to-cell communication, thereby providing a more comprehensive understanding of intercellular interactions within the network.
• Drug targets are identified based on their ability to control the GRN toward desired states, and the platform systematically predicts these through network control technology.
• Large-scale computer simulations further uncover the mechanisms of action for the identified drug targets.

Specification

MObyDiCK is a Python framework that supports coding in Jupyter Notebook, and coding environments can be provided on GCP Compute Engine instances upon customer request.

• Input: Single cell RNA sequencing data*, Desired (or undesired) state
• Output: Target gene list including first-in-class targets

<aside> 📢

*Data must be provided by the customer or sourced from publicly available datasets. If neither is available, a data generation package will be necessary.

</aside>

Deliverables

• Final reports within 8 weeks including:
  • Target gene lists
  • GRN
  • Mechanism analysis on GRN
  • In vitro validation results
• Progress meetings every 2 weeks.

Business Development

• Enterprise-level contract research
  • No IP share
  • NRE
  • Running royalty
• Co-development
  • IP share