ABC-Net: a divide-and-conquer based deep learning architecture for SMILES recognition from molecular images.

Structural information for chemical compounds is often described by pictorial images in most scientific documents, which cannot be easily understood and manipulated by computers. This dilemma makes optical chemical structure recognition (OCSR) an essential tool for automatically mining knowledge from an enormous amount of literature. However, existing OCSR methods fall far short of our expectations for realistic requirements due to their poor recovery accuracy. In this paper, we developed a deep neural network model named ABC-Net (Atom and Bond Center Network) to predict graph structures directly. Based on the divide-and-conquer principle, we propose to model an atom or a bond as a single point in the center. In this way, we can leverage a fully convolutional neural network (CNN) to generate a series of heat-maps to identify these points and predict relevant properties, such as atom types, atom charges, bond types and other properties. Thus, the molecular structure can be recovered by assembling the detected atoms and bonds. Our approach integrates all the detection and property prediction tasks into a single fully CNN, which is scalable and capable of processing molecular images quite efficiently. Experimental results demonstrate that our method could achieve a significant improvement in recognition performance compared with publicly available tools. The proposed method could be considered as a promising solution to OCSR problems and a starting point for the acquisition of molecular information in the literature.

DSGAT: predicting frequencies of drug side effects by graph attention networks.

A critical issue of drug risk-benefit evaluation is to determine the frequencies of drug side effects. Randomized controlled trail is the conventional method for obtaining the frequencies of side effects, while it is laborious and slow. Therefore, it is necessary to guide the trail by computational methods. Existing methods for predicting the frequencies of drug side effects focus on modeling drug-side effect interaction graph. The inherent disadvantage of these approaches is that their performance is closely linked to the density of interactions but which is highly sparse. More importantly, for a cold start drug that does not appear in the training data, such methods cannot learn the preference embedding of the drug because there is no link to the drug in the interaction graph. In this work, we propose a new method for predicting the frequencies of drug side effects, DSGAT, by using the drug molecular graph instead of the commonly used interaction graph. This leads to the ability to learn embeddings for cold start drugs with graph attention networks. The proposed novel loss function, i.e. weighted $\varepsilon$-insensitive loss function, could alleviate the sparsity problem. Experimental results on one benchmark dataset demonstrate that DSGAT yields significant improvement for cold start drugs and outperforms the state-of-the-art performance in the warm start scenario. Source code and datasets are available at https://github.com/xxy45/DSGAT.

DeepDDS: deep graph neural network with attention mechanism to predict synergistic drug combinations.

MOTIVATION: Drug combination therapy has become an increasingly promising method in the treatment of cancer. However, the number of possible drug combinations is so huge that it is hard to screen synergistic drug combinations through wet-lab experiments. Therefore, computational screening has become an important way to prioritize drug combinations. Graph neural network has recently shown remarkable performance in the prediction of compound-protein interactions, but it has not been applied to the screening of drug combinations. RESULTS: In this paper, we proposed a deep learning model based on graph neural network and attention mechanism to identify drug combinations that can effectively inhibit the viability of specific cancer cells. The feature embeddings of drug molecule structure and gene expression profiles were taken as input to multilayer feedforward neural network to identify the synergistic drug combinations. We compared DeepDDS (Deep Learning for Drug-Drug Synergy prediction) with classical machine learning methods and other deep learning-based methods on benchmark data set, and the leave-one-out experimental results showed that DeepDDS achieved better performance than competitive methods. Also, on an independent test set released by well-known pharmaceutical enterprise AstraZeneca, DeepDDS was superior to competitive methods by more than 16% predictive precision. Furthermore, we explored the interpretability of the graph attention network and found the correlation matrix of atomic features revealed important chemical substructures of drugs. We believed that DeepDDS is an effective tool that prioritized synergistic drug combinations for further wet-lab experiment validation. AVAILABILITY AND IMPLEMENTATION: Source code and data are available at https://github.com/Sinwang404/DeepDDS/tree/master.

Interactive Quantum Chemistry Enabled by Machine Learning, Graphical Processing Units, and Cloud Computing.

Modern quantum chemistry algorithms are increasingly able to accurately predict molecular properties that are useful for chemists in research and education. Despite this progress, performing such calculations is currently unattainable to the wider chemistry community, as they often require domain expertise, computer programming skills, and powerful computer hardware. In this review, we outline methods to eliminate these barriers using cutting-edge technologies. We discuss the ingredients needed to create accessible platforms that can compute quantum chemistry properties in real time, including graphical processing units-accelerated quantum chemistry in the cloud, artificial intelligence-driven natural molecule input methods, and extended reality visualization. We end by highlighting a series of exciting applications that assemble these components to create uniquely interactive platforms for computing and visualizing spectra, 3D structures, molecular orbitals, and many other chemical properties.

Toward integration of glycan chemical databases: an algorithm and software tool for extracting sugars from chemical structures.

Integration of glycan-related databases between different research fields is essential in glycoscience. It requires knowledge across the breadth of science because most glycans exist as glycoconjugates. On the other hand, especially between chemistry and biology, glycan data has not been easy to integrate due to the huge variety of glycan structure representations. We have developed WURCS (Web 3.0 Unique Representation of Carbohydrate Structures) as a notation for representing all glycan structures uniquely for the purpose of integrating data across scientific data resources. While the integration of glycan data in the field of biology has been greatly advanced, in the field of chemistry, progress has been hampered due to the lack of appropriate rules to extract sugars from chemical structures. Thus, we developed a unique algorithm to determine the range of structures allowed to be considered as sugars from the structural formulae of compounds, and we developed software to extract sugars in WURCS format according to this algorithm. In this manuscript, we show that our algorithm can extract sugars from glycoconjugate molecules represented at the molecular level and can distinguish them from other biomolecules, such as amino acids, nucleic acids, and lipids. Available as software, MolWURCS is freely available and downloadable ( https://gitlab.com/glycoinfo/molwurcs ).

Marine Delivery Vehicles: Molecular Components and Applications of Bacterial Extracellular Vesicles.

In marine ecosystems, communication among microorganisms is crucial since the distance is significant if considered on a microbial scale. One of the ways to reduce this gap is through the production of extracellular vesicles, which can transport molecules to guarantee nutrients to the cells. Marine bacteria release extracellular vesicles (EVs), small membrane-bound structures of 40 nm to 1 µm diameter, into their surrounding environment. The vesicles contain various cellular compounds, including lipids, proteins, nucleic acids, and glycans. EVs may contribute to dissolved organic carbon, thus facilitating heterotroph growth. This review will focus on marine bacterial EVs, analyzing their structure, composition, functions, and applications.

Relationships between the inhibitory efficacy and physicochemical properties of six organic acids and monolaurin against Bacillus weihenstephanensis KBAB4 growth in liquid medium.

Organic acids are widely used in foodstuffs to inhibit pathogen and spoiler growth. In this study, six organic acids (acetic, lactic, propionic, phenyllactic, caprylic, and lauric acid) and monolaurin were selected based on their physicochemical properties: their molecular structure (carbon chain length), their lipophilicity (logP), and their ability to dissociate in a liquid environment (pKa). The relation between these physicochemical properties and the inhibitory efficacy against B. weihenstephanensis KBAB4 growth was evaluated. After assessing the active form of these compounds against the strain (undissociated, dissociated or both forms), their MIC values were estimated in nutrient broth at pH 6.0 and 5.5 using two models (Lambert & Pearson, 2000; Luong, 1985). The use of two models highlighted the mode of action of an antibacterial compound in its environment, thanks to the additional estimation of the curve shape α or the Non-Inhibitory Concentration (NIC). The undissociated form of the tested acids is responsible for growth inhibition, except for lauric acid and monolaurin. Moreover, long-carbon chain acids have lower estimated MICs, compared to short-chain acids. Thus, the inhibitory efficacy of organic acids is strongly related to their carbon chain length and lipophilicity. Lipophilicity is the main mechanism of action of a membrane-active compound, it can be favored by long chain structure or high pKa in an acid environment like food.

The Effect of Methyl-Derivatives of Flavanone on MCP-1, MIP-1ß, RANTES, and Eotaxin Release by Activated RAW264.7 Macrophages.

Chemokines, also known as chemotactic cytokines, stimulate the migration of immune cells. These molecules play a key role in the pathogenesis of inflammation leading to atherosclerosis, neurodegenerative disorders, rheumatoid arthritis, insulin-resistant diabetes, and cancer. Moreover, they take part in inflammatory bowel disease (IBD). The main objective of our research was to determine the activity of methyl-derivatives of flavanone, namely, 2'-methylflavanone (5B), 3'-methylflavanone (6B), 4'-methylflavanone (7B), and 6-methylflavanone (8B), on the releasing of selected cytokines by RAW264.7 macrophages activated by LPS. We determined the concentration of chemokines belonging to the CC chemokine family, namely, MCP-1, MIP-1ß, RANTES, and eotaxin, using the Bio-Plex Magnetic Luminex Assay and the Bio-PlexTM 200 System. Among the tested compounds, only 5B and 6B had the strongest effect on inhibiting the examined chemokines' release by macrophages. Therefore, 5B and 6B appear to be potentially useful in the prevention of diseases associated with the inflammatory process.

Precisely Designed Morphology and Surface Chemical Structure of Fe-N-C Electrocatalysts for Enhanced Oxygen Reaction Reduction Activity.

Fe-N-C materials have been regarded as one of the potential candidates to replace traditional noble-metal-based electrocatalysts for the oxygen reduction reaction (ORR). It is believed that the structure of carbon support in Fe-N-C materials plays an essential role in highly efficient ORR. However, precisely designing the morphology and surface chemical structure of carbon support remains a challenge. Herein, we present a novel synthetic strategy for the preparation of porous carbon spheres (PCSs) with high specific surface area, well-defined pore structure, tunable morphology and controllable heteroatom doping. The synthesis involves Schiff-based polymerization utilizing octaaminophenyl polyhedral oligomeric silsesquioxane (POSS-NH2) and heteroatom-containing aldehydes, followed by pyrolysis and HF etching. The well-defined pore structure of PCS can provide the confinement field for ferroin and transform into Fe-N-C sites after carbonization. The tunable morphology of PCS can be easily achieved by changing the solvents. The surface chemical structure of PCS can be tailored by utilizing different heteroatom-containing aldehydes. After optimizing the structure of PCS, Fe-N-C loading on N,S-codoped porous carbon sphere (NSPCS-Fe) displays outstanding ORR activity in alkaline solution. This work paves a new path for fabrication of Fe-N-C materials with the desired morphology and well-designed surface chemical structure, demonstrating significant potential for energy-related applications.

Tailoring Ionic Liquid Chemical Structure for Enhanced Interfacial Engineering in Two-Step Perovskite Photovoltaics.

Ionic liquids (ILs) have emerged as versatile tools for interfacial engineering in perovskite photovoltaics. Their multifaceted application targets defect mitigation at SnO2 -perovskite interfaces, finely tuning energy level alignment, and enhancing charge transport, meanwhile suppressing non-radiative recombination. However, the diverse chemical structures of ILs present challenges in selecting suitable candidates for effective interfacial modification. This study adopted a systematic approach, manipulating IL chemical structures. Three ILs with distinct anions are introduced to modify perovskite/SnO2 interfaces to elevate the photovoltaic capabilities of perovskite devices. Specifically, ILs with different anions exhibited varied chemical interactions, leading to notable passivation effects, as confirmed by Density Functional Theory (DFT) calculation. A detailed analysis is also conducted on the relationship between the ILs' structure and regulation of energy level arrangement, work function, perovskite crystallization, interface stress, charge transfer, and device performance. By optimizing IL chemical structures and exploiting their multifunctional interface modification properties, the champion device achieved a PCE of 24.52% with attentional long-term stability. The study establishes a holistic link between IL structures and device performance, thereby promoting wider application of ILs in perovskite-based technologies.

PreDTIs: prediction of drug-target interactions based on multiple feature information using gradient boosting framework with data balancing and feature selection techniques.

Discovering drug-target (protein) interactions (DTIs) is of great significance for researching and developing novel drugs, having a tremendous advantage to pharmaceutical industries and patients. However, the prediction of DTIs using wet-lab experimental methods is generally expensive and time-consuming. Therefore, different machine learning-based methods have been developed for this purpose, but there are still substantial unknown interactions needed to discover. Furthermore, data imbalance and feature dimensionality problems are a critical challenge in drug-target datasets, which can decrease the classifier performances that have not been significantly addressed yet. This paper proposed a novel drug-target interaction prediction method called PreDTIs. First, the feature vectors of the protein sequence are extracted by the pseudo-position-specific scoring matrix (PsePSSM), dipeptide composition (DC) and pseudo amino acid composition (PseAAC); and the drug is encoded with MACCS substructure fingerings. Besides, we propose a FastUS algorithm to handle the class imbalance problem and also develop a MoIFS algorithm to remove the irrelevant and redundant features for getting the best optimal features. Finally, balanced and optimal features are provided to the LightGBM Classifier to identify DTIs, and the 5-fold CV validation test method was applied to evaluate the prediction ability of the proposed method. Prediction results indicate that the proposed model PreDTIs is significantly superior to other existing methods in predicting DTIs, and our model could be used to discover new drugs for unknown disorders or infections, such as for the coronavirus disease 2019 using existing drugs compounds and severe acute respiratory syndrome coronavirus 2 protein sequences.

Pharmacokinetics of the Recalcitrant Drug Lamotrigine: Identification and Distribution of Metabolites in Cucumber Plants.

Treated wastewater is an important source of water for irrigation. As a result, irrigated crops are chronically exposed to wastewater-derived pharmaceuticals, such as the anticonvulsant drug lamotrigine. Lamotrigine is known to be taken up by plants, but its plant-derived metabolites and their distribution in different plant organs are unknown. This study aimed to detect and identify metabolites of lamotrigine in cucumber plants grown for 35 days in a hydroponic solution by using LC-MS/MS (Orbitrap) analysis. Our data showed that 96% of the lamotrigine taken up was metabolized. Sixteen metabolites possessing a lamotrigine core structure were detected. Reference standards confirmed two; five were tentatively identified, and nine molecular formulas were assigned. The data suggest that lamotrigine is metabolized via N-carbamylation, N-glucosidation, N-alkylation, N-formylation, N-oxidation, and amidine hydrolysis. The metabolites LTG-N2-oxide, M284, M312, and M370 were most likely produced in the roots and were translocated to the leaves. Metabolites M272, M312, M314, M354, M368, M370, and M418 were dominant in leaves. Only a few metabolites were detected in the fruits. With an increasing exposure time, lamotrigine leaf concentrations decreased because of continuous metabolism. Our data showed that the metabolism of lamotrigine in a plant is fast and that a majority of metabolites are concentrated in the roots and leaves.

Machine Learning-Assisted Design of Thin-Film Composite Membranes for Solvent Recovery.

Organic solvents are extensively utilized in industries as raw materials, reaction media, and cleaning agents. It is crucial to efficiently recover solvents for environmental protection and sustainable manufacturing. Recently, organic solvent nanofiltration (OSN) has emerged as an energy-efficient membrane technology for solvent recovery; however, current OSN membranes are largely fabricated by trial-and-error methods. In this study, for the first time, we develop a machine learning (ML) approach to design new thin-film composite membranes for solvent recovery. The monomers used in interfacial polymerization, along with membrane, solvent and solute properties, are featurized to train ML models via gradient boosting regression. The ML models demonstrate high accuracy in predicting OSN performance including solvent permeance and solute rejection. Subsequently, 167 new membranes are designed from 40 monomers and their OSN performance is predicted by the ML models for common solvents (methanol, acetone, dimethylformamide, and n-hexane). New top-performing membranes are identified with methanol permeance superior to that of existing membranes. Particularly, nitrogen-containing heterocyclic monomers are found to enhance microporosity and contribute to higher permeance. Finally, one new membrane is experimentally synthesized and tested to validate the ML predictions. Based on the chemical structures of monomers, the ML approach developed here provides a bottom-up strategy toward the rational design of new membranes for high-performance solvent recovery and many other technologically important applications.

Recent progress in chemistry and bioactivity of monoterpenoid indole alkaloids from the genus gelsemium: a comprehensive review.

Monoterpenoid indole alkaloids (MIAs) represent a major class of active ingredients from the plants of the genus Gelsemium. Gelsemium MIAs with diverse chemical structures can be divided into six categories: gelsedine-, gelsemine-, humantenine-, koumine-, sarpagine- and yohimbane-type. Additionally, gelsemium MIAs exert a wide range of bioactivities, including anti-tumour, immunosuppression, anti-anxiety, analgesia, and so on. Owing to their fascinating structures and potent pharmaceutical properties, these gelsemium MIAs arouse significant organic chemists' interest to design state-of-the-art synthetic strategies for their total synthesis. In this review, we comprehensively summarised recently reported novel gelsemium MIAs, potential pharmacological activities of some active molecules, and total synthetic strategies covering the period from 2013 to 2022. It is expected that this study may open the window to timely illuminate and guide further study and development of gelsemium MIAs and their derivatives in clinical practice.

Structural Characterization and Anti-Osteoporosis Effects of a Novel Sialoglycopeptide from Tuna Eggs.

Several sialoglycopeptides were isolated from several fish eggs and exerted anti-osteoporosis effects. However, few papers have explored sialoglycopeptide from tuna eggs (T-ES). Here, a novel T-ES was prepared through extraction with KCl solution and subsequent enzymolysis. Pure T-ES was obtained through DEAE-Sepharose ion exchange chromatography and sephacryl S-300 gel filtration chromatography. The T-ES was composed of 14.07% protein, 73.54% hexose, and 8.28% Neu5Ac, with a molecular weight of 9481 Da. The backbone carbohydrate in the T-ES was →4)-ß-D-GlcN-(1→3)-α-D-GalN-(1→3)-ß-D-Glc-(1→2)-α-D-Gal-(1→2)-α-D-Gal-(1→3)-α-D-Man-(1→, with two branches of ß-D-GlcN-(1→ and α-D-GalN-(1→ linking at o-4 in →2,4)-α-D-Gal-(1→. Neu5Ac in the T-ES was linked to the branch of α-D-GlcN-(1→. A peptide chain, Ala-Asp-Asn-Lys-Ser*-Met-Ile that was connected to the carbohydrate chain through O-glycosylation at the -OH of serine. Furthermore, in vitro data revealed that T-ES could remarkably enhance bone density, bone biomechanical properties, and bone microstructure in SAMP mice. The T-ES elevated serum osteogenesis-related markers and reduced bone resorption-related markers in serum and urine. The present study's results demonstrated that T-ES, a novel sialoglycopeptide, showed significant anti-osteoporosis effects, which will accelerate the utilization of T-ES as an alternative marine drug or functional food for anti-osteoporosis.

Structural Characterization and Effects on Insulin Resistance of a Novel Chondroitin Sulfate from Halaelurus burgeri Skin.

Several studies have isolated chondroitin sulphate (CHS) from sharks' jaws or cartilage. However, there has been little research on CHS from shark skin. In the present study, we extracted a novel CHS from Halaelurus burgeri skin, which has a novel chemical structure and bioactivity on improvement in insulin resistance. Results using Fourier transform-infrared spectroscopy (FT-IR), 1H-nuclear magnetic resonance spectroscopy (1H-NMR), and methylation analysis showed that the structure of the CHS was [4)-ß-D-GlcpA-(1→3)-ß-D-GlcpNAc-(1→]n with 17.40% of sulfate group concentration. Its molecular weight was 238.35 kDa, and the yield was 17.81%. Experiments on animals showed that this CHS could dramatically decrease body weight, reduce blood glucose and insulin levels, lower lipid concentrations both in the serum and the liver, improve glucose tolerance and insulin sensitivity, and regulate serum-inflammatory factors. These results demonstrated that the CHS from H. burgeri skin has a positive effect in reducing insulin resistance because of its novel structure, which provides a significant implication for the polysaccharide as a functional food.

In Vitro and In Vivo Biocompatibility of Natural and Synthetic Pseudomonas aeruginosa Pyomelanin for Potential Biomedical Applications.

Bacteria are the source of many bioactive compounds, including polymers with various physiological functions and the potential for medical applications. Pyomelanin from Pseudomonas aeruginosa, a nonfermenting Gram-negative bacterium, is a black-brown negatively charged extracellular polymer of homogentisic acid produced during L-tyrosine catabolism. Due to its chemical properties and the presence of active functional groups, pyomelanin is a candidate for the development of new antioxidant, antimicrobial and immunomodulatory formulations. This work aimed to obtain bacterial water-soluble (Pyosol), water-insoluble (Pyoinsol) and synthetic (sPyo) pyomelanin variants and characterize their chemical structure, thermosensitivity and biosafety in vitro and in vivo (Galleria mallonella). FTIR analysis showed that aromatic ring connections in the polymer chains were dominant in Pyosol and sPyo, whereas Pyoinsol had fewer Car-Car links between rings. The differences in chemical structure influence the solubility of various forms of pyomelanins, their thermal stability and biological activity. Pyosol and Pyoinsol showed higher biological safety than sPyo. The obtained results qualify Pyosol and Pyoinsol for evaluation of their antimicrobial, immunomodulatory and proregenerative activities.

Chemical Structure Diversity and Extensive Biological Functions of Specialized Metabolites in Rice.

Rice (Oryza sativa L.) is thought to have been domesticated many times independently in China and India, and many modern cultivars are available. All rice tissues are rich in specialized metabolites (SPMs). To date, a total of 181 terpenoids, 199 phenolics, 41 alkaloids, and 26 other types of compounds have been detected in rice. Some volatile sesquiterpenoids released by rice are known to attract the natural enemies of rice herbivores, and play an indirect role in defense. Momilactone, phytocassane, and oryzalic acid are the most common diterpenoids found in rice, and are found at all growth stages. Indolamides, including serotonin, tryptamine, and N-benzoylserotonin, are the main rice alkaloids. The SPMs mainly exhibit defense functions with direct roles in resisting herbivory and pathogenic infections. In addition, phenolics are also important in indirect defense, and enhance wax deposition in leaves and promote the lignification of stems. Meanwhile, rice SPMs also have allelopathic effects and are crucial in the regulation of the relationships between different plants or between plants and microorganisms. In this study, we reviewed the various structures and functions of rice SPMs. This paper will provide useful information and methodological resources to inform the improvement of rice resistance and the promotion of the rice industry.

Structural and Serological Characterization of the O Antigen of Proteus mirabilis Clinical Isolates Classified into a New Proteus Serogroup, O84.

Two closely related Proteus mirabilis smooth strains, Kr1 and Ks20, were isolated from wound and skin samples, respectively, of two infected patients in central Poland. Serological tests, using the rabbit Kr1-specific antiserum, revealed that both strains presented the same O serotype. Their O antigens are unique among the Proteus O serotypes, which had been described earlier, as they were not recognized in an enzyme-linked immunosorbent assay (ELISA) by a set of Proteus O1-O83 antisera. Additionally, the Kr1 antiserum did not react with O1-O83 lipopolysaccharides (LPSs). The O-specific polysaccharide (OPS, O antigen) of P. mirabilis Kr1 was obtained via the mild acid degradation of the LPSs, and its structure was established via a chemical analysis and one- and two-dimensional 1H and 13C nuclear magnetic resonance (NMR) spectroscopy applied to both initial and O-deacetylated polysaccharides, where most ß-2-acetamido-2-deoxyglucose (N-acetylglucosamine) (GlcNAc) residues are non-stoichiometrically O-acetylated at positions 3, 4, and 6 or 3 and 6, and a minority of α-GlcNAc residues are 6-O-acetylated. Based on the serological features and chemical data, P. mirabilis Kr1 and Ks20 were proposed as candidates to a new successive O-serogroup in the genus Proteus, O84, which is another example of new Proteus O serotypes identified lately among serologically differentiated Proteus bacilli infecting patients in central Poland.

Preparation and Anti-Lung Cancer Activity Analysis of Guaiacyl-Type Dehydrogenation Polymer.

In this paper, guaiacyl dehydrogenated lignin polymer (G-DHP) was synthesized using coniferin as a substrate in the presence of ß-glucosidase and laccase. Carbon-13 nuclear magnetic resonance (13C-NMR) determination revealed that the structure of G-DHP was relatively similar to that of ginkgo milled wood lignin (MWL), with both containing ß-O-4, ß-5, ß-1, ß-ß, and 5-5 substructures. G-DHP fractions with different molecular weights were obtained by classification with different polar solvents. The bioactivity assay indicated that the ether-soluble fraction (DC2) showed the strongest inhibition of A549 lung cancer cells, with an IC50 of 181.46 ± 28.01 µg/mL. The DC2 fraction was further purified using medium-pressure liquid chromatography. Anti-cancer analysis revealed that the D4 and D5 compounds from DC2 had better anti-tumor activity, with IC50 values of 61.54 ± 17.10 µg/mL and 28.61 ± 8.52 µg/mL, respectively. Heating electrospray ionization tandem mass spectrometry (HESI-MS) results showed that both the D4 and D5 were ß-5-linked dimers of coniferyl aldehyde, and the 13C-NMR and 1H-NMR analyses confirmed the structure of the D5. Together, these results indicate that the presence of an aldehyde group on the side chain of the phenylpropane unit of G-DHP enhances its anticancer activity.