HyperLab
Service Introduction
How does adding AI to drug discovery research change the game?
Drug Discovery

AI Drug Discovery case study - EP1. Hyper Screening

We introduce a case study on utilizing Hyper Screening to develop a MARK4 inhibitor drug. This article covers the process from identifying initial hit compounds to analyzing experimental results.
medichem
Heejin Jung, R&D Strategy Team Lead
2025.01.207min read
AI Drug Discovery CASE STUDY - EP1. Hyper Screening.png

What You Will Learn in This Article

We introduce a case study on utilizing Hyper Screening to develop a MARK4 inhibitor drug. This article covers the process from identifying initial hit compounds to analyzing experimental results. By leveraging Hyper Screening, 16 hit compounds were identified from a pool of 107, and further cell viability tests led to the discovery of 1 compound with IC50 < 10 µM.

Who Should Read This Article?

If you have identified your target and associated disease for drug development but need to find initial hit compounds, this article demonstrates how Hyper Lab can help you get started effectively.

What is Hyper Lab?

Hyper Lab is an AI-powered drug discovery platform that accelerates and optimizes the drug development process. It offers tools for protein-ligand interaction prediction (Hyper Binding), novel molecular design (Hyper Design), drug property prediction (Hyper ADME/T), and large-scale molecular screening (Hyper Screening). This article focuses on how Hyper Screening was applied in this case study.

AI Drug Discovery Case Study: Target Preparation - MARK4

  • Defining the Role of MARK4 and Its Connection to Disease
    • MARK4 is a serine/threonine kinase that regulates the organization of microtubule arrays and plays a key role in maintaining cellular polarity.
  • Role of MARK Kinase in Disease
    • Phosphorylation of tau’s microtubule-binding domain by MARK kinase promotes hyperphosphorylation of tau by GSK-3 and CDK5, leading to the aggregation of tau into filaments and tangles. This process destabilizes and collapses microtubule networks. Inhibiting MARK4 is expected to suppress cancer cell proliferation, prevent metastasis, and induce apoptosis through specific signaling pathways.
    • Rationale for Target Selection
      • The relationship between MARK4 expression and cancer patient survival rates.
      • Elevated levels of MARK4L are associated with a worse prognosis.
      • Potential for first-in-class drug development.
      • The kinase characteristics of MARK4 provide abundant data for computational approaches.
      • Selection of an ATP competitive binding site.
  • Indication Setting

  • Analysis of the target protein's tissue distribution and its correlation with diseases led to the decision to focus on developing a treatment for glioma (Glioblastoma multiforme). The physicochemical requirements for glioma drug development were defined, which can be reviewed and filtered using Hyper Lab's ADME/T features.

  • Considering the therapeutic characteristics and cell permeability, the following physicochemical parameters were established:

    1. LogP: 3–5
    2. tPSA: <90
    3. Molecular Weight (MW): <500
    4. H-Bond Donors (HBD): <3
    5. Solubility: Moderate to Soluble

    Note: Hyper Lab's solubility standards predict log S (log [mol/L]) at neutral pH and classify it as follows:

    • Insoluble: log S < -6
    • Soluble: log S > -4
    • Moderate: -6 ≤ log S ≤ -4

AI Drug Discovery Case Study: MARK4 Target Analysis Using Hyper Lab

  1. After defining the target for drug research and selecting the associated disease, if you need to identify initial hit compounds, follow these steps to make the most of Hyper Lab.
    1. Use the Hyper Screening feature in Hyper Lab to perform virtual screening efficiently.
    2. Select and input the protein structure (PDB ID) into the platform. (Example: Hyper Lab interface screenshot)

2. Refer to relevant literature to identify and select the ATP binding site.

3. Since the goal is to develop an ATP competitive inhibitor, designate the hinge interaction as the primary contact residue:

    • Hinge: Ala 135 hydrogen bond interaction
    • After screening, compounds can be filtered based on the presence of hinge interactions.

4. Compile a list of reference compounds and input them into Hyper Lab, then set criteria based on the binding scores.

5. Perform Hyper Screening

    • Conduct virtual screening using either the Hyper Lab library or your in-house library.
    • We utilized the Diverse Library to perform virtual screening on one million compounds.

6. The screening results were processed within 24 hours, yielding a list of 500 compounds ranked by protein-compound binding energy. After applying ADMET property filters, 107 compounds were selected for purchase. (Vendor information and purchase links provided.)

7. Additionally, to ensure selectivity for MARK 1, 2, 3, and 4 isoforms, PDB data for each isoform was input, and binding scores were compared to reflect selectivity based on score differences.

AI Drug Discovery Case Study: Candidate Discovery Through Hyper Screening

  • Enzyme activity and cellular activity results successfully identified initial hit compounds.
  • Out of 107 compounds, 16 hit compounds were shortlisted, with 4 compounds showing performance on par with or better than the reference compound.
  • Further cell viability assays identified 1 compound with an IC50 < 10 µM.

  • Future Actions
    • Verify the novelty of active compounds to determine their intellectual property (IP) potential.
    • Optimize potency and refine in vitro ADME/T properties through structural modifications of the initial compounds.

AI Drug Discovery Case Study: Time and Cost Reduction with Hyper Screening

  • Time Savings
    • Traditional hit compound identification takes months and requires screening hundreds of compounds, but Hyper Screening completes this process within 24 hours.

  • Efficient Selection
    • Rapidly narrows down hundreds of compounds to a focused list of 500 promising candidates.
  • Web-Based Accessibility
    • Unlike conventional virtual screening, which often involves time-consuming collaborations with external companies, Hyper Screening allows users to select libraries directly on a web-based platform (custom library uploads supported). Acting as an AI-driven research assistant, Hyper Screening is a game-changing paradigm in drug development.

AI Drug Discovery Case Study: Sneak Peek of the Next Chapter

  • In the next article, we’ll discuss the structural optimization process for active compounds using Hyper Design.
  • If you have active compounds but require structural optimization, face challenges in selecting molecules without a dedicated chemist, or wish to learn structural modification techniques, stay tuned for the next article!

 

Free Trial: https://alie.kr/4mc2y7n

Contact us: https://abit.ly/6tr1uz

 

 

View List