Scholarlink

<p>Diffuse large B-cell lymphoma (DLBCL) remains difficult to treat due to relapse and resistance. Interleukin-1 receptor–associated kinase 4 (IRAK4) is a key driver of pro-survival signaling and represents a promising therapeutic target. Although several IRAK4 inhibitors have shown preclinical activity, their application in lymphoma remains limited. In this study, we applied an integrated computational strategy combining machine learning–based quantitative structure–activity relationship (QSAR) modelling, scaffold optimization, docking, absorption-distribution-metabolism-excretion-and-toxicity (ADMET) screening, and molecular dynamics to design novel IRAK4 inhibitors. The dataset, curated from ChEMBL and filtered to include only unique compounds with reported IC<sub>50</sub> values, provided a reliable foundation for model development. A dataset of over 2,000 compounds was reduced to 29 key descriptors, supporting a robust Random Forest model (R<sup>2</sup> = 0.94, Q<sup>2</sup> = 0.63, CCC = 0.97). Screening identified compound 9 as a lead scaffold with strong binding affinity (− 10.0&#xa0;kcal·mol⁻¹) and favourable pharmacokinetic predictions. Guided optimization produced analogue 9c, which showed enhanced activity, stable protein–ligand interactions in 100 ns simulations, and a molecular mechanics generalized Born surface area (MM-GBSA) binding free energy of − 77.2&#xa0;kcal·mol<sup>− 1</sup>. It is important to note that these results are computational predictions and await experimental validation. These findings nominate 9c as a promising preclinical candidate and highlight the value of combining data-driven prediction with structural refinement in lymphoma drug discovery.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

From scaffold to candidate: an integrated machine Learning, QSAR, docking, and molecular dynamics framework for the rational design of IRAK4 inhibitors in diffuse large B-cell lymphoma

  • 期刊论文

  • 发表日期：

    2025年12月27日

查看全文
  • Josiah Joseph Isah,
  • Adamu Uzairu,
  • Sani Uba,
  • Muhammad Tukur Ibrahim

摘要

Diffuse large B-cell lymphoma (DLBCL) remains difficult to treat due to relapse and resistance. Interleukin-1 receptor–associated kinase 4 (IRAK4) is a key driver of pro-survival signaling and represents a promising therapeutic target. Although several IRAK4 inhibitors have shown preclinical activity, their application in lymphoma remains limited. In this study, we applied an integrated computational strategy combining machine learning–based quantitative structure–activity relationship (QSAR) modelling, scaffold optimization, docking, absorption-distribution-metabolism-excretion-and-toxicity (ADMET) screening, and molecular dynamics to design novel IRAK4 inhibitors. The dataset, curated from ChEMBL and filtered to include only unique compounds with reported IC50 values, provided a reliable foundation for model development. A dataset of over 2,000 compounds was reduced to 29 key descriptors, supporting a robust Random Forest model (R2 = 0.94, Q2 = 0.63, CCC = 0.97). Screening identified compound 9 as a lead scaffold with strong binding affinity (− 10.0 kcal·mol⁻¹) and favourable pharmacokinetic predictions. Guided optimization produced analogue 9c, which showed enhanced activity, stable protein–ligand interactions in 100 ns simulations, and a molecular mechanics generalized Born surface area (MM-GBSA) binding free energy of − 77.2 kcal·mol− 1. It is important to note that these results are computational predictions and await experimental validation. These findings nominate 9c as a promising preclinical candidate and highlight the value of combining data-driven prediction with structural refinement in lymphoma drug discovery.