Central nervous system (CNS) diseases comprise a wide range of medical conditions that affect the spinal cord and brain. Neurodegenerative diseases (NDDs) represent a significant subset of CNS disorders characterized by the progressive degeneration of nerve cells. NDDs are age-associated multifactorial diseases characterized by dementia, cognitive impairment, memory decline, motor dysfunction, progressive loss of neuronal cells, and extensive brain damage (Baswar et al., 2021). Aging is recognized as a primary risk factor for the progression of neurodegeneration, which results in significant impairment in human wellbeing and health (Duc Nguyen, 2023). The incidence and prevalence of age-related neurodegeneration, which are predominantly observed in clinical practice, primarily encompass Alzheimer’s disease, Parkinson’s disease, and amyotrophic lateral sclerosis and are increasingly recognized as major causes of death and disability worldwide. The pathophysiological hallmarks among these NDDs are misfolded protein accumulation, oxidative stress, neuroinflammation, and mitochondrial dysfunction (Wang et al., 2010). The aging population around the world is contributing to the rise in the occurrence of neurological diseases, constituting a significant threat to healthcare systems. To address this problem, there is an urgent need to discover and develop novel neurotherapeutic strategies that can effectively target the underlying mechanisms of these neurological diseases (Duc Nguyen, 2023; Wang et al., 2010).

Neurotherapeutics represents a multidisciplinary field in neuroscience that focuses on developing novel treatments for neurological disorders. It comprises various applications such as therapeutic interventions, pharmacological agents, gene therapies, cognitive behavioral therapies, and neuromodulation techniques. These therapies seek to alleviate symptoms, prevent disease progression, and ultimately improve patient outcomes in various neurological conditions (Smith et al., 2012). Neurotherapeutic drug development aims to identify new treatments that will increase the quality of life of individuals suffering from neurologic disorders to NDDs. Progress in neurotherapeutics and neuropharmacology with respect to clinical methodology employing therapeutic agents was used to successfully treat both primary neurologic illnesses and neurodegenerative conditions (Cummings, 2006). Neurotherapeutic approaches have been utilized in NDD treatments of the CNS, resulting in the creation of biologically active molecules as medicines that particularly target underlying mechanisms involved in the disease etiology. However, the identification of active molecules has faced some difficulty in treating neurodegenerative conditions due to the presence of a specialized microvascular unit, known as the “blood–brain barrier” (BBB), that is crucial for the drug development process (Lawal et al., 2018).

The BBB is the complex network of brain microvessels that separates the CNS compartment from peripheral blood circulation. The primary cells that make up the BBB are brain microvascular endothelial cells, which are held together by neuronal cells such as astrocytes and pericytes that act according to the state of the CNS. Hence, the BBB plays a vital role in protecting the CNS by regulating the stability of the physiological (internal) environment of brain tissue, maintaining brain homeostasis and preserving neuronal viability (Abbott et al., 2010). The BBB facilitates the supply of essential nutrients necessary for the normal functioning of the brain. The highly selective nature of the BBB allows only specific molecules to pass through it and enter the brain, as it has the least permeable capillaries due to physical barriers (tight junctions). This has been considered as a major obstacle in designing and delivering beneficial drug-like compounds into the brain via the BBB to treat CNS diseases (Geldenhuys et al., 2015). However, the restrictive nature of the BBB allows only 2% of biologically active small molecules to cross the intact BBB to reach the brain at varying degrees (Mikitsh and Chacko, 2014). For many neurological diseases, there is almost no effective treatment due to the insufficient permeability of therapeutic agents into the brain through the BBB. Therefore, screening small molecules as to their BBB permeability is a prerequisite for a drug discovery process for treating CNS disorders and NDDs (Nagpal et al., 2013).

Neurological disorders necessitate prolonged and lifelong therapeutic interventions. Current therapeutic approaches predominantly rely on synthetic drugs that are commonly used to treat most NDDs which have adverse reactions and side effects (Feng et al., 2023). Natural products and dietary-based molecules have exhibited significant therapeutic efficacy in preventing major diseases. Additionally, natural resources provide a unique way for the identification of promising novel chemicals with validated efficacy. It is estimated that nearly 50% of all newly approved drugs can be traced back to a structural origin derived from a natural product (Zhang et al., 2017). There is a growing interest in exploring active molecules from diet or dietary interventions and functional foods that possess neuroprotective properties that may enhance their wellbeing and potentially slow down the progression of neurodegeneration. These approaches may provide neuroprotective benefits with fewer side effects than conventional pharmacological treatments. As research continues to evolve in this field, the integration of dietary interventions as complementary strategies with innovative drug-delivery approaches may facilitate the effective management of NDDs (Feng et al., 2023).

The development of neurotherapeutics is significantly impacted by accurate predictions of BBB permeability and CNS activity. Existing traditional methods of assessing BBB permeability are often very challenging, time-consuming, require large-scale experimental trials, and are laborious, resulting in low throughput (Shaker et al., 2023) and are hence unsuitable for screening large libraries of molecules. Therefore, given these experimental difficulties, there is a pressing need for innovative approaches that can facilitate the rapid screening of drug-like candidates for their ability to permeate the BBB (Könczöl et al., 2013). Recent advances in computational-aided drug development (CADD) have emerged as powerful techniques in drug discovery, offering the potential to screen active molecules and predict BBB permeability based on physicochemical and pharmacokinetic properties with higher predictability and clinical applicability (Aldewachi et al., 2021). Therefore, multiple computational tools, in silico techniques, and existing machine learning model-based approaches have been introduced to screen large libraries of compounds or small molecules that can quickly predict the BBB permeability of active compounds. Such predictive models are highly beneficial in enhancing the early phase of the drug discovery process, especially in the field of CNS research (Stéen et al., 2022). Accelerating the drug discovery process is essential for performing high-throughput virtual screening of drug-like small molecules to evaluate the BBB permeability of the active molecules.

This research applied in silico techniques together with machine learning and deep learning-based approaches in order to screen and predict the BBB permeability of druglike biologically active molecules from natural products and dietary sources. We explored the BBB permeability of CNS-active small molecules by interpreting physicochemical descriptors to compute the brain-to-blood ratio, as these metrics are routinely utilized to predict clinical exposure in CNS during the process of drug discovery (Zhang et al., 2016). Hence, we started with ligand-based virtual screening based on structural similarity to FDA-approved drugs. Then, the active molecules were screened and filtered based on multiple parameters, including BBB permeability, CNS activity, ADME profiling, toxicity, drug-likeness, bioavailability, medicinal chemistry, side effect resources, and bioactivity studies. Figure 1 provides a research workflow for screening BBB-permeable CNS-active molecules. All molecules were evaluated for the following parameters in order to identify drug-like small molecules: improved pharmacokinetics, pharmacodynamics, physicochemical properties, and the ability to directly activate at the CNS level by BBB permeation. Therefore, the incorporation of computational approaches in screening CNS-active and BBB-permeable molecules in the early phase of the drug discovery process can reduce the later-stage attrition rate and improve the overall success rate of drug discovery and development. This may enable the discovery of small molecules from natural sources with appropriate BBB permeation to elicit their bioactivity response against neurotherapeutic targets (Dichiara et al., 2024).

The blood–brain barrier (BBB) has become an important clinical feature in the discovery of small molecules that target neurodegenerative conditions (Sweeney et al., 2019). Effective delivery of neurotherapeutic agents, drugs, and small molecules across the BBB in order to reach the brain is a primary limiting factor for the successful therapy of neurological conditions. During the early stages of drug research, checking for BBB permeation offers crucial information for choosing the appropriate molecules. However, few studies have reported that many compounds failed due to their inability to permeate the BBB rather than a lack of potency to reach the brain, which rendered the BBB a significant obstacle in discovering central nervous system (CNS) drugs (Cornelissen et al., 2023). Therefore, our work aimed to address this challenge by employing in silico tools and existing machine learning model-based approaches to screen for BBB-permeable neurotherapeutics (Bicker et al., 2014), CNS-active drug-like small molecules that could serve as potential therapeutic interventions that can effectively target the underlying mechanisms of neurodegenerative diseases (NDDs) such as Alzheimer’s disease, Parkinson’s disease, and amyotrophic lateral sclerosis.

A total of 2,127 molecules were screened using a pharmacological model of five FDA-approved drugs. Ligand-based drug discovery requires a target-site-specific drug in order to understand the chemical space and create novel drug-like molecules (Bung et al., 2022). Hence, we performed high-throughput virtual screening based on structural similarity with respect to the selected list of FDA-approved drugs for NDDs, such as Donepezil, tacrine, memantine, galantamine, and rivastigmine. These pharmacotherapies appear to partially alleviate various clinical symptoms associated with NDDs. However, the therapeutic benefits are often minimal, transient, and non-selective, and are frequently accompanied by adverse side effects (Duc Nguyen, 2023; Chopade et al., 2023). Due to these drawbacks, innovative therapies from natural products and dietary sources are being considered for NDDs. A wide range of molecular descriptors explores trends with CNS active drugs. The involvement of 12 types of molecular descriptors in our study proved to be associated with BBB permeability of existing studies (Jiang et al., 2016; Faramarz et al., 2022) and provided a comprehensive understanding of how these features correlate with drug-like properties. Specific physicochemical descriptors like MW, lipophilicity, TPSA, and HB can influence BBB permeability (log BB) and are positively correlated to brain uptake (Faramarz et al., 2022; Vilar et al., 2010). The druggability of each molecule must adhere to a defined set of standard rules like RO5 to favor BBB permeation and reach the brain via passive diffusion (Pajouhesh and Lenz, 2005). Therefore, molecules that do not obey these criteria are unsuitable for clinical use and are susceptible to various challenges regarding their properties associated with ADME.

BBB-permeable molecules can have varying effects on the BBB structure. Conversely, certain molecules have been shown to restore or maintain BBB integrity by upregulating tight junction proteins and inhibiting inflammatory mediators. Compounds, such as those derived from natural products and dietary sources, have demonstrated both high BBB permeability and neuroprotective effects, with minimal toxicity to the BBB (Sánchez-Martínez et al., 2022a; Sánchez-Martínez et al., 2022b). According to the BBB model of Shaker et al. (2021), and Tang et al. (2022), 582 small molecules were identified as BBB+ with over ∼85% accuracy in differentiating BBB+ and BBB− based on the brain-to-blood ratio (Zhang et al., 2017). BBB-permeable molecules have been found with direct CNS activities based on predicted logBB values, which is crucial for therapeutic efficacy (Garg et al., 2008). A total of six key descriptors were taken into account for CNS multi-parameter optimization. The CNS MPO showed that 75% of the FDA-approved drugs for CNS disorders have high desirability scores ≥4 for testing molecules, correlating well with our screened molecules for better translation into the clinics (Wager et al., 2016). Medicinal chemists and neuroscientists in recent years optimized the identified molecules to obey essential criteria of synthetic accessibility and lead-likeness in order to improve pharmacokinetics and pharmacodynamics, has been reported to enhance brain uptake via BBB (Anthony et al., 2021; Sanghai et al., 2024).

Physicochemical parameters and structural properties of small molecules determine the BBB permeability, particularly in relation to passive diffusion mechanisms (Cornelissen et al., 2023). We found that the oral bioavailability of small molecules with predicted values exceeding the threshold of 30% indicates good absorption potential in the gastrointestinal tract, which is particularly important for brain targeting. Lipophilicity and molecular weight were found to be pivotal in determining bioavailability, aligned with ADME parameters: increased passive permeability, minimal P-gp liability, adequate metabolic stability, long half-lives, and efficient clearance rates (Stéen et al., 2022). In compliance with the existing study, it was reported that an absorption of drugs more than 90% of the administered dose leads to increased BBB permeability. According to ProTox (Banerjee et al., 2024), a virtual lab assessed the safety profile of desired molecules and found them to be non-toxic. In addition, it is reported to be inactive/less active against organ toxicity, toxicity endpoints, and toxicological pathways (Tox21). Adverse drug reactions were assessed in order to avoid unwanted CNS side effects at the site of action (Adenot and Lahana, 2004). The drug-likeness determines whether the druggability of each small molecule that evolved from natural products is BBB-permeable due to lipid solubility, has a high degree of hydrogen bonding, positively charges molecules, a low Mw of 400–600 Da, and the highest TPSA value (Aldewachi et al., 2021; Russo et al., 2018; Cornelissen et al., 2023). Stéen et al. (2022) suggested that the hydrogen bond interactions between a molecule and the hydrophilic portion of the BBB through lipophilicity enable brain uptake.

Neuronal repair processes involve axonal regeneration, synaptic reorganization, and neuroprotection. Molecules that promote these processes are of great interest for treating NDDs and CNS injuries with neuroregenerative potential. Notably, certain flavonoids and terpenoids have been reported to support neuronal repair and neuroprotection. Neuronal repair depends on the neurobiological activity of small molecules based on their bioavailability in the brain (Wang et al., 2022). Emerging evidence indicates that small molecules possess multimodal biological activities, including antioxidant, anti-inflammatory, anticholinergic, and kinase-inhibitory effects. This multifunctionality is increasingly recognized as a significant characteristic of small molecules, enabling them to interact with multiple biological targets and pathways. Activity spectra predict nearly 4000 types of bioactivities based on the structure–activity relationship (Filimonov et al., 2014). Moreover, our findings indicate that small molecules permeating the BBB possess various neuroprotective activities, such as nootropic, BDNF enhancer, and the modulation of neuroinflammatory pathways. These findings align with current therapeutic needs in treating neurodegenerative diseases. However, the existing scientific research provides extensive lists and classes of such compounds, and many have been confirmed to cross the BBB and exert neuroprotective effects (Mohd Sairazi and Sirajudeen, 2020).

The 112 active CNS molecules identified in this study comprise a diverse array of naturally based small molecules, each with demonstrated BBB permeability. Recent research highlights the distinct mechanisms of action across key brain cell types, such as neurons, astrocytes, microglia, and endothelial cells, for effective therapeutic intervention in NDD. Active molecules that reach the brain can regulate CNS homeostasis, reduce neuroinflammation and excitotoxicity, promote neuroprotection by reducing oxidative stress, inhibit apoptosis, and support synaptic plasticity, thus directly modulating neurotransmitter systems, including dopaminergic, glutamatergic, and cholinergic pathways, or promoting neuronal survival, making them relevant for treating conditions such as Parkinson’s disease, Alzheimer’s disease, and amyotrophic lateral sclerosis (Cornelissen et al., 2023; Kocsis et al., 2025; Grabska-Kobyłecka et al., 2023).

The chemical diversity of CNS-active, BBB-permeable molecules, including flavonoids, terpenoids, polyphenols, and alkaloids, are well-documented in natural product research as promising scaffolds for CNS drug development (Isabel et al., 2024). The inclusion of naturally derived compounds is consistent with recent reviews that emphasize the BBB-crossing potential of various natural products (Isabel et al., 2024). Our findings are promising and largely support the ability of specific dietary molecules to promote CNS health, improve neuroprotective mechanisms, enhance cognitive function, and mitigate the risk of NDDs. Furthermore, our results align with existing studies that suggest a correlation between dietary intake and neurological outcomes. Many studies have used multiple ligands and/or small molecules in a single delivery system to maximize brain uptake and targeted delivery with improved BBB permeation (Cornelissen et al., 2023; Isabel et al., 2024; Al Gailani et al., 2022). In future, experimental validation needs to be carried out by ensuring their impact on BBB integrity and the neuroprotective properties of small molecules, which can then be formulated as valuable food additives, food supplements, and nutraceuticals that could provide new insights into neurotherapeutics as promising therapeutic interventions for NDDs.

Neurodegenerative diseases will continue to rise with an aging population. Nevertheless, the rate of advancement in neurotherapeutics is limited by the complexities associated with drug delivery across the blood–brain barrier (BBB). Thus, the integration of in silico techniques and machine learning models can provide greater insights to accurately screen and identify biologically active molecules that can effectively pass through the BBB. A total of 112 effective BBB-permeable neurotherapeutic molecules were identified after screening a library of small molecules. Potential small molecules were then configured for pharmacokinetic properties and pharmacodynamic characteristics for recommendation as primary lead molecules. These active molecules were prioritized by considering important factors like structural similarity, ADME profiling, toxicity endpoints, drug-likeness, bioavailability criteria, and side effects. Resources were applied individually for the evaluation of safety profiles of drug-like small molecules to recognize possible health risks. Moreover, bioactivity studies relevant to neuronal activity revealed that many of these molecules possess the capacity to modulate critical pathways involved in neuroprotection and neuroregeneration. Medicinal chemistry friendliness ensured that only those molecules with optimal drug-like characteristics were made for potential therapeutic use, with improved pharmacokinetic properties, expected central nervous system (CNS) activity, and permeation phenomena at the BBB level. While our results are promising, it is significant to note that the translation of in silico predictions into successful clinical applications remains challenging. The complexity of the BBB and the multifactorial nature of neurodegenerative diseases (NDDs) necessitate experimental validation of the identified molecules, as well as their ability to restore BBB integrity and provide neuroprotection. Ultimately, the overall research was aimed at the development of novel neurotherapeutics delivered across the BBB with CNS activity. Additionally, our research underscores the potential for dietary interventions as a strategy to promote brain health and mitigate the risk of NDDs like dementia/Alzheimer’s disease, Parkinson’s disease, amyotrophic lateral sclerosis, and multiple sclerosis. Future research should focus on experimental validation by elucidating the underlying mechanisms and optimizing dietary formulations to maximize their beneficial effects on the CNS.