Flavonoids from Daphniphyllum calycinum as candidate ferroptosis inducers in lung adenocarcinoma: an integrated network pharmacology, molecular dynamics and multi-omics study
Description
This is an integrative computational pharmacology study that starts with plant flavonoids and applies them to the context of ferroptosis in lung adenocarcinoma; the original clinical data are sourced from TCGA-LUAD and GSE131907, while the rest primarily consists of database annotations and computational results generated in this study.
Files
Steps to reproduce
This study did not involve new cell, animal, or human experiments, and therefore did not involve experimental instruments or reagents. Data acquisition was divided into two categories. The first category consists of publicly available database data: using "Niu’erfeng" and "Daphniphyllum calycinum" as search terms, data were retrieved from TCMSP in July 2023 unless otherwise specified, combined with published phytochemical literature to collect active components; screening was performed using OB ≥ 30% and DL ≥ 0.18, compound structures were obtained from PubChem, and human targets were predicted via SwissTargetPrediction, retaining targets with a prediction probability > 0.1. Lung adenocarcinoma-related genes were retrieved from GeneCards, OMIM, and DisGeNET, and ferroptosis-related genes were obtained from FerrDb V2; gene names were standardized according to HGNC symbols. Protein-protein interaction networks were constructed using STRING v11 with a confidence threshold > 0.4 and analyzed with Cytoscape 3.9.1 and CytoHubba. Functional enrichment analysis was performed using DAVID 2021. RNA-seq data of TCGA-LUAD were obtained from UCSC Xena, and clinical data were obtained from cBioPortal, including 515 tumor samples, 59 adjacent normal samples, and 503 survival analysis samples, analyzed using R 4.3.0 and edgeR v3.42. Single-cell data were obtained from GEO dataset GSE131907, which included 208,506 cells from 44 patients. The second category consists of computational data generated in this study: target protein structures were obtained from RCSB PDB, ligand structures from PubChem, and structural preparation and molecular docking were performed using PyMOL, Open Babel v3.1.1, and AutoDock Vina v1.2.5; the luteolin–BCL2 complex was selected for 100 ns molecular dynamics simulation using GROMACS with the AMBER14SB force field and TIP3P water model, and binding free energy was calculated using gmx_MMPBSA. Immune infiltration analysis was performed using TIMER2.0, TISIDB, and ImmuCellAI, with TIDE used only for exploratory assessment. All database versions, retrieval dates, screening thresholds, software versions, and analysis procedures are detailed in the Methods section.