Neurodevelopmental disorders (NDDs) represent significant public health challenges due to their multifactorial etiology and clinical heterogeneity. Current treatments remain limited, highlighting the need for novel therapeutic strategies. This study aimed to identify neuroprotective natural compounds targeting NDD-associated pathways and describe an integrative computational pipeline combining in silico screening, network pharmacology, and molecular docking approaches to accelerate NDD drug discovery. An integrative computational pipeline was developed through sequential phases: (1) systematic screening of the Traditional Chinese Medicine Systems Pharmacology Database (TCMSP) for natural compounds meeting drug-likeness criteria and toxicity thresholds; (2) biological activity prediction; (3) network pharmacology analysis integrating compound targets and NDD-associated genes; (4) protein–protein interaction network construction and functional enrichment; and (5) molecular docking validation of top compounds against prioritized targets. From 2634 initial compounds, 10 met all selection criteria. Network analysis revealed significant interactions between compound targets and NDD-associated genes, with enrichment in neurodevelopment, cognition, and synaptic regulation pathways. Three key targets emerged as hubs: CSNK2B, GRIN1, and MAPK1. Molecular docking demonstrated high-affinity binding of caryophyllene oxide, linoleic acid, and tangeretin, supported by stable interactions with catalytic residues. This study identifies caryophyllene oxide, linoleic acid, and tangeretin as promising multi-target compounds for NDD intervention, with verified interactions against key neurodevelopmental targets. The integrative computational pipeline effectively bridges traditional medicine knowledge with modern drug discovery, offering a strategy to accelerate neurotherapeutic development while reducing experimental costs. These findings warrant further experimental validation of the prioritized compounds.
Neurodevelopmental disorders (NDDs) represent a heterogeneous group of conditions originating from atypical alterations in brain development, primarily occurring during the prenatal stages and typically becoming apparent in childhood or adolescence. Although phenotypically variable, these disorders exhibit core symptomatic overlaps, including cognitive deficits, motor dysfunction, language impairments, and emotional regulation difficulties.
The Diagnostic and Statistical Manual of Mental Disorders (DSM-5) classifies NDDs as encompassing Intellectual Disability, Communication Disorders, Autism Spectrum Disorder (ASD), Attention-Deficit/Hyperactivity Disorder (ADHD), Neurodevelopmental Motor Disorders, and Specific Learning Disorders. While demonstrating strong heritability, these conditions emerge through complex gene–environment interactions that influence neurodevelopmental trajectories.
Evaluating genetic risk in NDDs is challenging due to their genetic heterogeneity. ASD exemplifies this complexity, with hundreds of implicated risk genes showing minimal overlap between affected individuals. Compounding this genetic variability, adverse environmental exposures interact with susceptibility loci to disrupt typical neurodevelopment, underscoring the critical role of gene–environment interplay in the pathophysiological mechanisms of NDDs.
NDDs pose an escalating global health challenge, significantly compromising quality of life and creating substantial socioeconomic impacts on healthcare systems. Although pharmacological treatments have advanced, the need for novel therapeutic approaches remains evident. The clinical heterogeneity of NDDs and the intricate gene–environment dynamics render single-gene targeting strategies particularly challenging and with limited therapeutic scope. Consequently, pathway-based approaches that modulate core neurodevelopmental mechanisms may offer more robust and comprehensive treatment alternatives.
Building on these pathway-based therapeutic opportunities, mechanistic studies consistently highlight three pathophysiological domains: (I) protein synthesis, (II) epigenetic transcriptional regulation, and (III) synaptic signaling. In ASD specifically, dysregulated pathways with therapeutic potential include Wnt/β-catenin, Sonic Hedgehog (Shh), ERK/MAPK, and PI3K/AKT signaling, along with neuroinflammatory mediators (TGFs and JAK/STAT) and neurotransmitter systems including glutamatergic and GABAergic imbalance. This pathway-level understanding enables targeted therapeutic development that addresses the multifactorial nature of these disorders.
Drug development is a lengthy and costly process, often marked by high failure rates, frequently due to efficacy or safety issues identified during clinical trials. To address these challenges, computational approaches have been increasingly adopted to streamline early discovery phases and improve candidate selection. Collectively termed Computer-Aided Drug Design (CADD), these methodologies leverage in silico techniques to prioritize compounds with optimal pharmacological profiles.
CADD encompasses two primary approaches: Structure-Based Drug Design (SBDD) and Ligand-Based Drug Design (LBDD). SBDD utilizes computational techniques including molecular docking and molecular dynamics simulations to predict ligand–target interactions. LBDD employs Quantitative Structure–Activity Relationship (QSAR) and pharmacophore modeling to identify novel bioactive compounds based on established structure–activity patterns. Furthermore, CADD integrates predictive pharmacokinetic and toxicological profiling before experimental testing, thereby reducing costs and risks in drug development.
Network pharmacology represents another powerful computational drug discovery methodology that adopts a systems perspective. This paradigm shift, first formalized by Hopkins, analyzes drug actions through interconnected target–pathway networks rather than single targets, enabling both drug repositioning and novel compound discovery, through polypharmacology mechanisms. The methodology has demonstrated broad therapeutic utility, with successful applications spanning cardiovascular (pulmonary hypertension), metabolic (diabetes mellitus), neurodegenerative (Alzheimer’s disease), and oncological (cancer) disorders.
The pursuit for innovative treatments for neurodevelopmental disorders has increasingly focused on natural compounds exhibiting neuroprotective properties. Notably, several of these phytochemicals exhibit antioxidant, anti-inflammatory, and neurotransmitter pathway-modulating activities, positioning them as candidates for addressing the complex pathophysiology of NDDs.
Among these neuroprotective phytochemicals, luteolin demonstrates CNS immunomodulation by inhibiting immune cell activation in the central nervous system and suppressing the expression of pro-inflammatory mediators such as IL-6 and TNF-α, in addition to modulating the NF-κB signaling pathway. Resveratrol exerts neuroprotective effects by stimulating mitochondrial biogenesis via the SIRT1/PGC-1α pathway, thereby reducing oxidative stress and neuroinflammation. Furthermore, curcumin exhibits antioxidant and anti-inflammatory properties, improving intracellular glutathione levels, reducing the production of reactive oxygen species (ROS), and protecting against mitochondrial damage, in addition to modulating inflammatory factors involved in the pathophysiology of neurodevelopmental disorders.
Given the demonstrated ability of natural compounds to modulate key biological pathways in NDDs, there is a pressing need to systematically identify novel neuroprotective molecules. While current pharmacotherapy primarily addresses symptoms (e.g., stimulants for ADHD, risperidone for irritability in ASD), natural compounds offer potential disease-modifying effects by targeting underlying mechanisms. The discovery of such molecules could enable the development of (1) complementary approaches to enhance existing therapies, (2) multi-target strategies addressing NDD complexity, and (3) personalized interventions based on individual pathway dysregulation.
This study aims to identify novel neuroprotective natural compounds for NDDs through an integrated computational approach combining in silico screening and network pharmacology. The innovative methodology employs a phased screening strategy that systematically evaluates pharmacokinetic properties, toxicity profiles, biological activity predictions, and mechanism-of-action analyses.
This multi-parametric approach enables more accurate drug candidate selection, optimizing the identification of promising compounds while significantly reducing the time and costs associated with early-stage drug development. By integrating critical pharmacological parameters into a unified pipeline, our strategy overcomes limitations of conventional drug discovery methods, offering an efficient pathway for developing multi-target therapies directed at core NDD pathophysiological mechanisms.
The natural compounds were systematically evaluated for key pharmacological properties, including drug-likeness and blood–brain barrier (BBB) permeability potential, followed by comprehensive toxicity endpoint and biological activity predictions. Through this multi-stage screening pipeline, ten compounds demonstrating favorable pharmacological profiles were selected for further investigation. The final candidates comprised six terpenoids [kobusone, caryophyllene oxide, α-humulene epoxide, isokobusone, selina-4(14),7(11)-dien-8-one, and miltionone II], one coumarin (osthol), two flavonoids (tangeretin and sinensetin), and one fatty acid (linoleic acid).
An initial pool of 2634 natural compounds was retrieved from the TCMSP database using ‘cognitive deficits’ as the primary search criterion. Sequential filtering through established drug-likeness parameters (TCMSP and SwissADME criteria) and removal of duplicate entries yielded 460 qualified candidates. The rigorous filtration process reduced the initial dataset by 82.5%, focusing on compounds with clinically translatable potential for cognitive enhancement. The parameter values evaluated for the ten selected natural compounds are presented in Table 1.
The canonical SMILES representations of the 460 compounds were submitted to the ProTox 3.0 platform.
