Fingerprint Similarity Research Articles

Combining crystallographic and binding affinity data towards a novel dataset of small molecule overlays

Although small molecule superposition is a standard technique in drug discovery, a rigorous performance assessment of the corresponding methods is currently challenging. Datasets in this field are sparse, small, tailored to specific applications, unavailable, or outdated. The newly developed LOBSTER set described herein offers a publicly available and method-independent dataset for benchmarking and method optimization. LOBSTER stands for “Ligand Overlays from Binding SiTe Ensemble Representatives”. All ligands were derived from the PDB in a fully automated workflow, including a ligand efficiency filter. So-called ligand ensembles were assembled by aligning identical binding sites. Thus, the ligands within the ensembles are superimposed according to their experimentally determined binding orientation and conformation. Overall, 671 representative ligand ensembles comprise 3583 ligands from 3521 proteins. Altogether, 72,734 ligand pairs based on the ensembles were grouped into ten distinct subsets based on their volume overlap, for the benefit of introducing different degrees of difficulty for evaluating superposition methods. Statistics on the physicochemical properties of the compounds indicate that the dataset represents drug-like compounds. Consensus Diversity Plots show predominantly high Bemis–Murcko scaffold diversity and low median MACCS fingerprint similarity for each ensemble. An analysis of the underlying protein classes further demonstrates the heterogeneity within our dataset. The LOBSTER set offers a variety of applications like benchmarking multiple as well as pairwise alignments, generating training and test sets, for example based on time splits, or empirical software performance evaluation studies. The LOBSTER set is publicly available at https://doi.org/10.5281/zenodo.12658320, representing a stable and versioned data resource. The Python scripts are available at https://github.com/rareylab/LOBSTER, open-source, and allow for updating or recreating superposition sets with different data sources. Graphical abstractSimplified illustration of the LOBSTER dataset generation.

Exploring the Potential of Adaptive, Local Machine Learning in Comparison to the Prediction Performance of Global Models: A Case Study from Bayer's Caco-2 Permeability Database.

Machine learning (ML) techniques are being widely implemented to fill the gap in simple molecular design guidelines for newer therapeutic modalities in the extended and beyond rule of five chemical space (eRo5, bRo5). These ML techniques predict molecular properties directly from the structure, allowing for the prioritization of promising compounds. However, the performance of models varies greatly among ML use cases. A molecular property for which achieving sufficient performance in generalizing global models still remains difficult is Caco-2 permeability. Especially within the lower permeability ranges, which are specific for larger molecules belonging to the e/bRo5 space, accurate regression predictions have proven to be challenging. The present study, therefore, identifies a suitable combination of ML algorithm and descriptors, consisting of the LightGBM algorithm and RDKit molecular property descriptors, to predict Caco-2 permeability very efficiently by a simple global model. An additionally introduced local model uses the same algorithm and descriptors but selects its training data based on Tanimoto fingerprint similarity to match the individual test compound's structure. Evaluation of this adaptive model, by systematically varying the number of most similar structures for training, shows that, in comparison to the global model, there was only marginally improved performance with specific training data constellations. These random improvements indicate that deriving general rules for local model parametrization is not possible a priori for the chosen algorithm and descriptor combination, and preselecting training data does not seem advantageous over global ML based on all available data, while creation of more data-efficient models was generally proven to be possible.

Similarity Analysis of Computer-Generated and Commercial Libraries for Targeted Biocompatible Coded Amino Acid Replacement.

Many non-natural amino acids can be incorporated by biological systems into coded functional peptides and proteins. For such incorporations to be effective, they must not only be compatible with the desired function but also evade various biochemical error-checking mechanisms. The underlying molecular mechanisms are complex, and this problem has been approached previously largely by expert perception of isomer compatibility, followed by empirical study. However, the number of amino acids that might be incorporable by the biological coding machinery may be too large to survey efficiently using such an intuitive approach. We introduce here a workflow for searching real and computed non-natural amino acid libraries for biosimilar amino acids which may be incorporable into coded proteins with minimal unintended disturbance of function. This workflow was also applied to molecules which have been previously benchmarked for their compatibility with the biological translation apparatus, as well as commercial catalogs. We report the results of scoring their contents based on fingerprint similarity via Tanimoto coefficients. These similarity scoring methods reveal candidate amino acids which could be substitutable into modern proteins. Our analysis discovers some already-implemented substitutions, but also suggests many novel ones.

Improved indoor localization using k‐medoids and k‐nearest neighbour algorithms with context similarity coefficient‐based fingerprint similarity metric

AbstractFingerprint database clustering and localization using k‐medoids and k‐nearest neighbour (k‐NN) algorithms respectively typically use distance‐based fingerprint similarity metrics, with their performances dependent on the type of distance metric used. This paper proposes employing a pattern‐based metric, the context similarity coefficient (CSC), for both algorithms instead of traditional distance‐based metrics. The CSC accounts for fingerprint behaviour and the non‐linear relationships among fingerprints during the similarity measurement. The performance of both algorithms with the CSC as the similarity metric is evaluated on four publicly available fingerprint databases, using position root mean square error (RMSE) and silhouette score as performance metrics. These results are compared to those of the same algorithms using five distance‐based metrics: Euclidean, square Euclidean, Manhattan, cosine, and Chebyshev distances. The k‐medoids algorithm with CSC shows moderate clustering performance compared to the five distance‐based metrics considered. However, when combined with the k‐NN algorithm also using CSC, it achieves the highest localization accuracy, with at least a 29% improvement in position RMSE across all four databases. The results indicate that while k‐medoids with CSC may not create well‐separated clusters, combining it with the k‐NN algorithm with CSC as its similarity metric significantly enhances localization accuracy compared to distance‐based metrics.

A rapid fingerprint positioning method based on deep convolutional neural network for MIMO-OFDM systems

The combination of fingerprint positioning and 5G (the 5th Generation Mobile Communication Technology) offers broader application prospects for indoor positioning technology, but also brings challenges in real-time performance. In this paper, we propose a fingerprint positioning method based on a deep convolutional neural network (DCNN) using a classification approach in a single-base station scenario for massive multiple input multiple output-orthogonal frequency division multiplexing (MIMO-OFDM) systems. We introduce an angle-delay domain fingerprint matrix that simplifies the computation process and increases the location differentiation. The cosine distance is chosen as the fingerprint similarity criterion due to its sensitivity to angular differences. First, the DCNN model is used to determine the sub-area to which the mobile terminal belongs, and then the weighted K-nearest neighbor (WKNN) matching algorithm is used to estimate the position within the sub-area. The positioning performance is simulated in a DeepMIMO indoor environment, showing that the classification DCNN method reduces the positioning time by 77.05% compared to the non-classification method, with only a 1.08% increase in average positioning error.

Exploring the role of Yuxuebi tablet in neuropathic pain with the method of similarity research of drug pharmacological effects based on unsupervised machine learning.

Having multiple pharmacological effects is a characteristic of Traditional Chinese Medicine (TCM). Currently, there is a lack of suitable methods to explore and discover modern diseases suitable for TCM treatment using this characteristic. Unsupervised machine learning technology is an efficient strategy to predict the pharmacological activity of drugs. This study takes Yuxuebi Tablet (YXB) as the research object. Using the unsupervised machine learning technology of drug cell functional fingerprint similarity research, the potential pharmacological effects of YXB were discovered and verified. LC-MS combined with the in vitro intestinal absorption method was used to identify components of YXB that could be absorbed by the intestinal tract of rats. Unsupervised learning hierarchical clustering was used to calculate the degree of similarity of cellular functional fingerprints between these components and 121 marketed Western drugs whose indications are diseases and symptoms that YXB is commonly used to treat. Then, based on the Library of Integrated Network-based Cellular Signatures database, pathway analysis was performed for selected Western drugs with high similarity in cellular functional fingerprints with the components of YXB to discover the potential pharmacological effects of YXB, which were validated by animal experiments. We identified 40 intestinally absorbed components of YXB. Through predictive studies, we found that they have pharmacological effects very similar to non-steroidal anti-inflammatory drugs (NSAIDs) and corticosteroids. In addition, we found that they have very similar pharmacological effects to anti-neuropathic pain medications (such as gabapentin, duloxetine, and pethidine) and may inhibit the NF-κB signaling pathway and biological processes related to pain perception. Therefore, YXB may have an antinociceptive effect on neuropathic pain. Finally, we demonstrated that YXB significantly reduced neuropathic pain in a rat model of sciatic nerve chronic constriction injury (CCI). Transcriptome analysis further revealed that YXB regulates the expression of multiple genes involved in nerve injury repair, signal transduction, ion channels, and inflammatory response, with key regulatory targets including Sgk1, Sst, Isl1, and Shh. This study successfully identified and confirmed the previously unknown pharmacological activity of YXB against neuropathic pain through unsupervised learning prediction and experimental verification.

Enhancing protein-ligand binding affinity prediction through sequential fusion of graph and convolutional neural networks.

Predicting protein-ligand binding affinity is a crucial and challenging task in structure-based drug discovery. With the accumulation of complex structures and binding affinity data, various machine-learning scoring functions, particularly those based on deep learning, have been developed for this task, exhibiting superiority over their traditional counterparts. A fusion model sequentially connecting a graph neural network (GNN) and a convolutional neural network (CNN) to predict protein-ligand binding affinity is proposed in this work. In this model, the intermediate outputs of the GNN layers, as supplementary descriptors of atomic chemical environments at different levels, are concatenated with the input features of CNN. The model demonstrates a noticeable improvement in performance on CASF-2016 benchmark compared to its constituent CNN models. The generalization ability of the model is evaluated by setting a series of thresholds for ligand extended-connectivity fingerprint similarity or protein sequence similarity between the training and test sets. Masking experiment reveals that model can capture key interaction regions. Furthermore, the fusion model is applied to a virtual screening task for a novel target, PI5P4Kα. The fusion strategy significantly improves the ability of the constituent CNN model to identify active compounds. This work offers a novel approach to enhancing the accuracy of deep learning models in predicting binding affinity through fusion strategies.

Medical device similarity analysis: a promising approach to medical device equivalence regulation

ABSTRACT Background This study aims to facilitate the identification of similar devices for both, the European Medical Device Regulation (MDR) and the US 510(k) equivalence pathway by leveraging existing data. Both are related to the regulatory pathway of read across for chemicals, where toxicological data from a known substance is transferred to one under investigation, as they aim to streamline the accreditation process for new devices and chemicals. Research design and methods This study employs latent semantic analysis to generate similarity values, harnessing the US Food and Drug Administration 510k-database, utilizing their ‘Device Descriptions’ and ‘Intended Use’ statements. Results For the representative inhaler cluster, similarity values up to 0.999 were generated for devices within a 510(k)-predicate tree, whereas values up to 0.124 were gathered for devices outside this group. Conclusion Traditionally, MDR equivalence involves manual review of many devices, which is laborious. However, our results suggest that the automated calculation of similarity coefficients streamlines this process, thus reducing regulatory effort, which can be beneficial for patients needing medical devices. Although this study is focused on the European perspective, it can find application within 510(k) equivalence regulation. The conceptual approach is reminiscent of chemical fingerprint similarity analysis employed in read-across.

Discovery of new antimicrobial thiophene derivatives with activity against drug-resistant Gram negative-bacteria.

Our aim is to identify new small molecules with antimicrobial potential, especially against colistin-resistant (Col-R) Acinetobacter baumannii and Escherichia coli. After initial hits identification by fingerprint similarity, MIC of 24 heterocyclic derivatives for A. baumannii and E. coli reference strains, and bactericidal activity of selected thiophenes against Col-R strains were determined. We analyzed changes in bacterial membrane permeability and the OMPs profile. Additionally, we determined bacterial adherence to host cells and performed molecular docking studies to assess their binding to bacterial targets. The compounds' MICs ranged from 4 to >64mg/L. Thiophene derivatives 4, 5 and 8 exhibited MIC50 values between 16 and 32mg/L for Col-R A. baumannii and 8 and 32mg/L for Col-R E. coli. The time-kill curve assay demonstrated that thiophenes 4 and 8 had bactericidal effects against Col-R A. baumannii and E. coli. Furthermore, treatment with them resulted in increased membrane permeabilization and reduced adherence of these isolates to host cells. Finally, the docking studies showed a stronger binding affinity to CarO1 and Omp33 of A. baumannii and OmpW and OmpC of E. coli. These findings indicate that thiophene derivatives possess antibacterial activity against Col-R A. baumannii and E. coli, suggesting that they may enhance the repertoire of drug treatments against bacteria.

Optimization and Spectrum-Effect Analysis of Ultrasonically Extracted Antioxidant Flavonoids from Persicae Ramulus.

The objectives of this study were to optimize the ultrasonic-assisted flavonoid extraction process from PR and to establish fingerprints in order to analyze the spectrum-effect relationship of antioxidant activity. The ultrasonic-assisted flavonoid extraction process from PR was optimized using RSM, and the fingerprints of twenty-eight batches of flavonoids from PR were established using UHPLC. Meanwhile, the in vitro antioxidant activity of PR was evaluated in DPPH and ABTS free radical-scavenging experiments. Then, the peaks of the effective antioxidant components were screened using the spectrum-effect relationships. The results show that the optimal extraction yield of flavonoids from PR was 3.24 ± 0.01 mg/g when using 53% ethanol, a 1:26 (g/mL) solid-liquid ratio, and 60 min of ultrasonic extraction. Additionally, the clearance of two antioxidant indices by the flavonoids extracted from PR had different degrees of correlation and showed concentration dependence. Simultaneously, the similarity of the UHPLC fingerprints of twenty-eight batches of PR samples ranged from 0.801 to 0.949, and four characteristic peaks, namely peaks 4, 12, 21, and 24, were screened as the peaks of the components responsible for the antioxidant effect of PR using a GRA, a Pearson correlation analysis, and a PLS-DA. In this study, characteristic peaks of the antioxidant effects of PR were screened in an investigation of the spectrum-effect relationship to provide a scientific basis for the study of pharmacodynamic substances and the elucidation of the mechanism of action of the antioxidant effect of PR.

Establishment of an HPLC Fingerprint and Cluster Analysis for Miao Ethnic Medicine Osbeckia opipara

: Osbeckia opipara, a traditional Miao medicine, is commonly used by the renowned national-level Chinese Traditional Medicine practitioner Zhengshi Wu for the treatment of diarrhea due to its strong antioxidative, anti-inflammatory, and antidiarrheal effects. This study aimed to establish a high-performance liquid chromatography (HPLC) fingerprint for O. opipara to provide new evidence and technical means for the scientific evaluation and effective quality control of O. opipara. The procedure involved isolation with a Nano ChromCore C18 column (250 mm × 4.6 mm, 5 μm), using a gradient elution of 0.1% formic acid in water and 0.1% formic acid in acetonitrile as the mobile phase, with a flow rate of 1.0 mL/min, a column temperature of 30°C, an injection volume of 10 μL, and detection at a wavelength of 254 nm. Under these chromatographic conditions, fingerprint analysis was conducted on 11 batches of O. opipara collected from different origins. The National Pharmacopoeia Committee developed the 'Chromatographic Fingerprint Similarity Evaluation System' (2004A version) for automated comparison, similarity computation, and analysis of chromatographic data. The results revealed 13 common peaks across the 11 batches of O. opipara samples, with a similarity to the automatically generated reference spectrum exceeding 0.9. SPSS 26.0 software was used to conduct cluster analysis on the peak areas of the 13 common peaks. The observations indicated that the reference spectrum generated from the 11 batches could serve as the standard fingerprint profile for O. opipara, providing sufficient characteristic information extraction.

Fully Flexible Molecular Alignment Enables Accurate Ligand Structure Modeling.

Accurate protein-ligand binding poses are the prerequisites of structure-based binding affinity prediction and provide the structural basis for in-depth lead optimization in small molecule drug design. However, it is challenging to provide reasonable predictions of binding poses for different molecules due to the complexity and diversity of the chemical space of small molecules. Similarity-based molecular alignment techniques can effectively narrow the search range, as structurally similar molecules are likely to have similar binding modes, with higher similarity usually correlated to higher success rates. However, molecular similarity is not consistently high because molecules often require changes to achieve specific purposes, leading to reduced alignment precision. To address this issue, we propose a new alignment method─Z-align. This method uses topological structural information as a criterion for evaluating similarity, reducing the reliance on molecular fingerprint similarity. Our method has achieved success rates significantly higher than those of other methods at moderate levels of similarity. Additionally, our approach can comprehensively and flexibly optimize bond lengths and angles of molecules, maintaining a high accuracy even when dealing with larger molecules. Consequently, our proposed solution helps in achieving more accurate binding poses in protein-ligand docking problems, facilitating the development of small molecule drugs. Z-align is freely available as a web server at https://cloud.zelixir.com/zalign/home.

Do Molecular Fingerprints Identify Diverse Active Drugs in Large-Scale Virtual Screening? (No).

Computational approaches for small-molecule drug discovery now regularly scale to the consideration of libraries containing billions of candidate small molecules. One promising approach to increased the speed of evaluating billion-molecule libraries is to develop succinct representations of each molecule that enable the rapid identification of molecules with similar properties. Molecular fingerprints are thought to provide a mechanism for producing such representations. Here, we explore the utility of commonly used fingerprints in the context of predicting similar molecular activity. We show that fingerprint similarity provides little discriminative power between active and inactive molecules for a target protein based on a known active-while they may sometimes provide some enrichment for active molecules in a drug screen, a screened data set will still be dominated by inactive molecules. We also demonstrate that high-similarity actives appear to share a scaffold with the query active, meaning that they could more easily be identified by structural enumeration. Furthermore, even when limited to only active molecules, fingerprint similarity values do not correlate with compound potency. In sum, these results highlight the need for a new wave of molecular representations that will improve the capacity to detect biologically active molecules based on their similarity to other such molecules.

Elucidation of anti-pneumonia pharmacodynamic material basis and potential mechanisms of Xiebai San by combining spectrum–efficacy relationship and surface plasmon resonance

Ethnopharmacological relevanceXiebai San (XBS), a classic Chinese prescription, has been used for the clinical treatment of pneumonia-related diseases for thousands of years. However, the anti-pneumonia pharmacodynamic material basis of XBS and its underlying mechanisms remain unclear. Aim of the studyThis study aimed to comprehensively investigate and verify the anti-pneumonia pharmacodynamic material basis and mechanisms of XBS. Materials and methodsThis study explored the anti-pneumonia activity and key pneumonia targets of XBS in lipopolysaccharide (LPS)-induced zebrafish and RAW264.7 cells in vivo and in vitro through transcriptomics, western blotting, and reverse transcription–quantitative polymerase chain reaction (RT–qPCR). The chemical fingerprint of XBS was established using high-performance liquid chromatography, and the similarities and areas of characteristic peaks of 15 batches of XBS were analyzed. Based on the spectrum–efficacy relationship, the potential anti-inflammatory components were screened according to their peak areas and efficacy using principal component analysis (PCA), bivariate correlation, and partial least squares regression analysis. Active components that bind to core targets were further screened based on surface plasmon resonance (SPR). The binding mode of proteins and components was simulated via molecular docking, which enabled the identification of the primary active components of XBS, thereby elucidating its anti-pneumonia properties. Finally, the anti-inflammatory activities of these components were verified in vitro. ResultsXBS decreased neutrophil aggregation in zebrafish and nitric oxide (NO) secretion in RAW264.7 cells as well as suppressed the release of downstream inflammatory cytokines such as iNOS, TNF-α, IL-1β, IL-18, and CXCL10 related to TNF and JAK–STAT signaling pathways. The phosphorylation of IκBα, Akt, and Stat3 was alleviated after XBS in cells. The fingerprint similarities of 15 batches of XBS ranged from 0.381 to 0.994, with a large difference. A total of 15 characteristic peaks were identified, and the relative standard deviation of their peak areas ranged from 24.1% to 70.7%. The results of in vitro anti-inflammatory activities of 15 batches of XBS showed that all samples inhibited the expression levels of NO and nine inflammatory markers. The anti-inflammatory index of 15 batches of XBS was determined to be 0.69–0.96 based on transformation of the anti-inflammatory rate and composite index method via PCA. The spectrum–efficacy relationship model of 15 characteristic peak areas and the anti-inflammatory index showed that 7 main potential active components were related to the anti-inflammatory activity of XBS. Moreover, four components (mulberroside A, isoquercitrin, liquiritigenin, and glycyrrhizic acid) screened based on SPR had different affinities toward TNFR1, Akt1, and Stat3 proteins, and the binding modes were elucidated via molecular docking. Finally, in LPS-induced RAW264.7 cells, all four active components (at a concentration of 60 μM) significantly inhibited the expression levels of NO and inflammatory markers. ConclusionsBased on the comprehensive strategy of spectrum–efficacy relationship and SPR, mulberroside A, isoquercitrin, liquiritigenin, and glycyrrhizic acid were identified as the primary pharmacodynamic active components involved in the anti-pneumonia activity of XBS and were found to intervene in TNF and JAK–STAT signaling pathways.

Optimization of the extraction process for Shenshou Taiyi powder based on Box-Behnken experimental design, standard relation, and FAHP-CRITIC methods

BackgroundAncient classic prescription play a crucial role in the preservation and advancement of traditional Chinese medicine (TCM) theories. They represent a significant milestone in the ongoing development and transmission of TCM knowledge and practices and are considered one of the breakthroughs in the development of TCM inheritance. In the process of developing ancient classic prescriptions, many problems may still arise in ensuring quality consistency between traditional methods and modern production processes, among which the extraction process poses major challenges. This paper introduces a practical approach extracting an ancient classic prescription using a modern extraction process. The technique is demonstrated through the study of the extraction process of Shenshou Taiyi powder (STP).MethodsThis study focuses on optimising the STP extraction process to ensure consistency in the quality of the product obtained through ancient and modern processes using the standard relation and fuzzy analytic hierarchical process (FAHP) and criteria importance through intercriteria correlation (CRITIC) method integrated weights combined with the Box-Behnken response surface test. Using the contents of rosmarinic acid, isoimperatorin, puerarin, as well as the extract yield and fingerprint similarity as evaluation indexes of STP, the Box-Behnken response surface method was employed to examine the varying extraction parameters, including water addition ratio, extraction duration, and number of extractions. The weighted coefficients for each parameter were calculated by combining the benchmark correlation and FAHP-CRITIC method, deriving a comprehensive score.ResultsThe optimal extraction process for STP consisted of a two extractions, each using at a tenfold quantity of water, performed for one hour. Process verification across three separate batches yielded a comprehensive score of 94.7, with a relative standard deviation of 0.76%.ConclusionsThe application of the Box-Behnken response surface method combined with standard relation and FAHP-CRITIC approach proved to be stable and feasible for optimising the extraction process of STP.

Identification of Reaction Network Hypotheses for Complex Feedstocks from Spectroscopic Measurements with Minimal Human Intervention.

In this work, we detail an automated reaction network hypothesis generation protocol for processes involving complex feedstocks where information about the species and reactions involved is unknown. Our methodology is process agnostic and can be utilized in any reactive process with spectroscopic measurements that provide information on the evolution of the components in the mixture. We decompose the mixture spectra to obtain spectroscopic signatures of the individual components and use a 1-D convolutional neural network to automatically identify functional groups indicated by them. We employ atom-atom mapping to automatically recover reaction rules that are applied on candidate molecules identified from chemistry databases through fingerprint similarity. The method is tested on synthetic data and on spectroscopic measurements of lab-scale batch hydrothermal liquefaction (HTL) of biomass to determine the accuracy of prediction across datasets of varying complexities. Our methodology is able to identify reaction network hypotheses containing reaction networks close to the ground truth in the case of synthetic data, and we are also able to recover candidate molecules and reaction networks close to the ones reported in the previous literature studies for biomass pyrolysis.

Fingerprint Correlation Between Twins

CODY BARNES, KRISTY HENSON, MARK FLOOD, Department of Natural Sciences, Fairmont State University, Fairmont, WV, 26201 Fingerprint correlation between identical twins When teaching about the uniqueness of fingerprints it is commonly said that identical twin fingerprint characteristics are not identical however, limited research has been conducted to determine if there are any correlations between similar print characteristics. In the recent past unrelated individuals have been misidentified based on fingerprint similarities in forensics cases and this may be more likely among twins. This research case study’s objective is to determine if there are any correlations between fingerprint characteristics of identical twins and the probability of misidentification. Identical twin fingerprints were rolled and collected onto 10 print cards then blindly analyzed using Henry Classification, AFIS, and NCIC techniques. Preliminary results indicate a strong relationship between characteristics, but it is not significant (0.49; p = 0.35); fingerprint characteristics between hands were not significantly different (R: p = 0.75, L: p = 0.24). Results indicate there could be a possibility of misidentification, but more twin fingerprints should be analyzed.

Computer-Assisted Drug Discovery of Potential African Anti-SARS-CoV-2 Natural Products Targeting the Helicase Protein

Objectives: In our continuous efforts to combat COVID-19, our objective was to conduct a comprehensive computer-aided drug design study utilizing 4924 African natural metabolites sourced from diverse databases across various African regions from 1962 to 2019. The primary goal was to target the SARS-CoV-2 helicase, a crucial enzyme in viral replication. Methods: We employed structural fingerprint and molecular similarity studies with VXG, the co-crystallized ligand, as a reference. Subsequently, docking and absorption, distribution, metabolism, excretion, and toxicity (ADMET) studies were conducted. Results: The structural fingerprint analysis identified 200 structurally similar compounds, then the molecular similarity studies selected 40 compounds. Among them, 15 metabolites with low free energies, favorable binding modes, and promising ADMET properties were identified. Four compounds were excluded according to the toxicity studies. Compound 1552, 1-(( S)-2,3-dihydro-2-(( R)-1-hydroxypropan-2-yl)benzofuran-5-yl)ethenone, exhibited the most favorable docking results. Molecular dynamics simulations conclusively demonstrated its stable binding to the SARS-CoV-2 helicase, characterized by low energy and optimal dynamics. Conclusion: The findings suggest promising avenues for potential COVID-19 cures, encouraging further exploration through in vitro and in vivo studies of the identified compounds, particularly compound 1552.

Identification of Vernonia patula Merr. and Its Similar Varieties Based on a Combination of HPLC Fingerprinting and Chemical Pattern Recognition.

Vernonia patula Merr. (VP) is a traditional medicine used by the Zhuang and Yao people, known for its therapeutic properties in treating anemopyretic cold and other diseases. Distinguishing VP from similar varieties such as Praxelis clematidea (PC), Ageratum conyzoides L. (AC) and Ageratum houstonianum Mill (AH) was challenging due to their similar traits and plant morphology. The HPLC fingerprints of 40 batches of VP and three similar varieties were established. SPSS 20.0 and SIMCA-P 13.0 were used to statistically analyze the chromatographic peak areas of 37 components. The results showed that the similarity of the HPLC fingerprints for each of the four varieties was >0.9, while the similarity between the control chromatogram of VP and its similar varieties was <0.678. Cluster analysis and partial least squares discriminant analysis provided consistent results, indicating that all four varieties could be individually clustered together. Through further analysis, we found isochlorogenic acid A and isochlorogenic acid C were present only in the original VP, while preconene II was present in the three similar varieties of VP. These three components are expected to be identification points for accurately distinguishing VP from PC, AC and AH.

Mix-Key: graph mixup with key structures for molecular property prediction.

Molecular property prediction faces the challenge of limited labeled data as it necessitates a series of specialized experiments to annotate target molecules. Data augmentation techniques can effectively address the issue of data scarcity. In recent years, Mixup has achieved significant success in traditional domains such as image processing. However, its application in molecular property prediction is relatively limited due to the irregular, non-Euclidean nature of graphs and the fact that minor variations in molecular structures can lead to alterations in their properties. To address these challenges, we propose a novel data augmentation method called Mix-Key tailored for molecular property prediction. Mix-Key aims to capture crucial features of molecular graphs, focusing separately on the molecular scaffolds and functional groups. By generating isomers that are relatively invariant to the scaffolds or functional groups, we effectively preserve the core information of molecules. Additionally, to capture interactive information between the scaffolds and functional groups while ensuring correlation between the original and augmented graphs, we introduce molecular fingerprint similarity and node similarity. Through these steps, Mix-Key determines the mixup ratio between the original graph and two isomers, thus generating more informative augmented molecular graphs. We extensively validate our approach on molecular datasets of different scales with several Graph Neural Network architectures. The results demonstrate that Mix-Key consistently outperforms other data augmentation methods in enhancing molecular property prediction on several datasets.