HyperLab
Service Introduction
An AI technology that makes the world's
medicines faster?
Drug Discovery

How HyperLab Helped Discover Real Drug Candidates

This article highlights two case studies showing how the AI drug discovery platform HyperLab effectively identified lead compounds and optimized drug candidates in real-world research.
작은 히츠 로고
HITS HyperLab Team
2025.08.0812min read
제목 (25) (1).png

HyperLab: An AI-Driven Drug Development Platform Powering Real-World Research

In the early stages of drug development, time and accuracy are paramount. Identifying effective compounds for a specific target within a limited timeframe directly impacts research efficiency. This is where HyperLab, an AI-driven drug development platform, stands out.

In this article, we introduce two case studies from published papers that demonstrate how HyperLab was used to optimize derivatives and identify effective compounds in real-world research. Let’s explore how researchers leveraged HyperLab to achieve their desired outcomes.

Optimizing Anticancer Derivatives with HyperLab: The Discovery of 7MeERT (Biomolecules 2024)

This study aimed to develop a novel anticancer drug effective against various cancer cells, including brain tumors, EGFR-overexpressing cancers, and multiple myeloma. The target protein was NDUFA12, and while the existing drug Ertredin interacts with this protein, its anticancer activity was limited in certain cancer cells.

Using HITS’s AI-driven drug development platform HyperLab (https://hyperlab.hits.ai/, as of July 12, 2024), researchers analyzed the interaction mechanisms of 7MeERT and Ertredin with the NDUFA12 protein, along with their physicochemical properties, including ADME/T predictions (PDB ID: O97725.pdb). Additionally, binding energies were calculated to evaluate how strongly each compound binds to NDUFA12 (Biomolecules 2024).

To address this, the research team utilized HyperLab to perform molecular docking and AI-driven hybrid virtual screening. As a result, they predicted that 7MeERT, a new derivative with a methyl group introduced to Ertredin, would exhibit superior binding affinity to NDUFA12. Additionally, HyperLab’s ADME/T prediction and structure-activity relationship (SAR) analysis features were used to evaluate drug-likeness and optimization strategies.

Experimental results showed that 7MeERT demonstrated stronger binding affinity than Ertredin, with a lower binding energy (−3.7 vs. −3.3 kcal/mol, where lower negative values indicate stronger binding). It also exhibited superior antiproliferative effects and apoptosis induction capabilities.

This case demonstrates how HyperLab’s AI predictions enabled early candidate selection, significantly reducing the cost and time required for drug development.

Source: Biomolecules 2024 paper (https://www.mdpi.com/2218-273X/14/9/1197)

Designing LATS Inhibitors with HyperLab: A Case Study in Experimental Validation of AI Analysis (ResearchGate 2024)

The paper “Discovery of Selective LATS Inhibitors via Scaffold Hopping,” published last year, represents a successful application of HyperLab’s Hyper Binding function in real-world research. In this study, researchers utilized HyperLab’s Hyper Binding function to identify candidate compounds with high activity against the target protein.

The research team first analyzed the key interactions between the protein and the ligand based on the binding structure of the known reference compound, ‘Truli.’ Subsequently, they leveraged HyperLab’s Hyper Binding function to predict the binding structures and affinities of new molecules, evaluating whether the core interactions were maintained.

Table 1. Synthesized structures and the percent enzymatic activities.

As a result, candidate compounds 5k and 5l, which exhibited high binding scores and maintained core interactions, were identified. Furthermore, using HyperLab’s 2D and 3D structure viewer functions, researchers confirmed that the hydrogen bonding interactions with Met154 and Glu156 in the ligand structure were preserved, consistent with the reference compound.

Fig. 3. Docking simulation and predicted human metabolic stabilities of 1, 5k and 5l.

In conclusion, candidate compounds predicted by HyperLab to have high activity demonstrated excellent activity in experiments. This case validates that considering both binding energy and binding structure significantly enhances the success rate of Hyper Binding-based analysis.

Source: ResearchGate 2024 paper ("Discovery of selective LATS inhibitors via scaffold hopping")

Accelerating Drug Development Possibilities with HyperLab’s AI

This article has explored how HyperLab’s Hyper Binding and Hyper ADME/T functions were applied in real-world research through two published papers. From predicting binding affinities between target proteins and molecules to analyzing binding structures via 2D and 3D viewers, HyperLab offers a range of features to support the early stages of drug development.

  • Hyper Screening: Identifies the top 500 molecules with high binding affinity to a target protein from a pool of 1 million molecules in just one day.
  • Hyper Screening X: Screens over 11 trillion compounds to select synthesizable, effective molecules in up to 48 hours.
  • Hyper Design: Enhances molecular structures with AI to simultaneously achieve novelty and activity.

Beyond these, HyperLab’s intuitive UI/UX ensures accessibility for all users. However, new users may find the setup or analysis process challenging. To assist, we offer free consulting services. For those looking to integrate AI into their drug development process, we recommend scheduling a consultation through the inquiry link below. As validated by published research, we are confident that HyperLab can significantly enhance the efficiency and accuracy of drug development.

AI-Powered Drug Discovery Platform: HyperLab

View List