Rethinking the applicability domain analysis in QSAR models.

Notwithstanding the wide adoption of the OECD principles (or best practices) for QSAR modeling, disparities between in silico predictions and experimental results are frequent, suggesting that model predictions are often too optimistic. Of these OECD principles, the applicability domain (AD) estimation has been recognized in several reports in the literature to be one of the most challenging, implying that the actual reliability measures of model predictions are often unreliable. Applying tree-based error analysis workflows on 5 QSAR models reported in the literature and available in the QsarDB repository, i.e., androgen receptor bioactivity (agonists, antagonists, and binders, respectively) and membrane permeability (highest membrane permeability and the intrinsic permeability), we demonstrate that predictions erroneously tagged as reliable (AD prediction errors) overwhelmingly correspond to instances in subspaces (cohorts) with the highest prediction error rates, highlighting the inhomogeneity of the AD space. In this sense, we call for more stringent AD analysis guidelines which require the incorporation of model error analysis schemes, to provide critical insight on the reliability of underlying AD algorithms. Additionally, any selected AD method should be rigorously validated to demonstrate its suitability for the model space over which it is applied. These steps will ultimately contribute to more accurate estimations of the reliability of model predictions. Finally, error analysis may also be useful in "rational" model refinement in that data expansion efforts and model retraining are focused on cohorts with the highest error rates.

Toxicity of selected pharmaceuticals and their mixtures to the aquatic indicators Daphnia magna and Aliivibrio fischeri.

Despite the benefits derived from the use of pharmaceuticals, these compounds are currently considered contaminants of emerging concern because of their presence and persistence in the environment. This study aimed to determine the toxicity of 27 pharmaceuticals and the interaction effects of binary mixtures of selected compounds towards two model organisms: the microcrustacean Daphnia magna and the bacterium Aliivibrio fischeri (Microtox test). Six compounds, namely polymyxin B, polymyxin E, fluoxetine, diphenhydramine, clenbuterol and ketoprofen exhibited moderate toxicity towards D. magna. Additionally, three compounds (cefotaxime, polymyxin B, polymyxin E) also showed a moderate toxic effect on A. fischeri. The comparison of such results with model estimations showed inaccuracy in the predicted data, highlighting the relevance of experimental ecotoxicological assays. The assayed mixtures contained four selected drugs of high-hazard according to their reported concentrations in wastewater and surface water (diphenhydramine, trimethoprim, ketoprofen, and fluoxetine); data revealed interactions only in the fluoxetine-containing mixtures for D. magna, while all mixtures showed interactions (mostly synergistic) for Microtox. Chronic effects on the reproduction of D. magna were observed after exposure to fluoxetine and diphenhydramine, although higher sensitivity was determined for the latter, while the mixture of these compounds (which showed acute synergy in both models) also affected the reproduction patterns. Nonetheless, all the effects described at the acute or chronic level (for individual compounds or mixtures) were determined at concentrations higher than commonly reported at environmental levels. This work provides valuable ecotoxicological information for the risk assessment of pharmaceuticals and their mixtures in the environment.

ElectroPredictor: An Application to Predict Mayr's Electrophilicity E through Implementation of an Ensemble Model Based on Machine Learning Algorithms.

Electrophilicity (E) is one of the most important parameters to understand the reactivity of an organic molecule. Although the theoretical electrophilicity index (ω) has been associated with E in a small homologous series, the use of w to predict E in a structurally heterogeneous set of compounds is not a trivial task. In this study, a robust ensemble model is created using Mayr's database of reactivity parameters. A combination of topological and quantum mechanical descriptors and different machine learning algorithms are employed for the model's development. The predictability of the model is assessed using different statistical parameters, and its validation is examined, including a training/test partition, an applicability domain, and a y-scrambling test. The global ensemble model presents a Q5-fold2 of 0.909 and a Qext2 of 0.912, demonstrating an excellent predictability performance of E values and showing that w is not a good descriptor for the prediction of E, especially for the case of neutral compounds. ElectroPredictor, a noncommercial Python application (https://github.com/mmoreno1/ElectroPredictor), is developed to predict E. QM9, a well-known large dataset containing 133885 neutral molecules, is used to perform a virtual screening (94.0% coverage). Finally, the 10 most electrophilic molecules are analyzed as possible new Mayr's electrophiles, which have not yet been experimentally tested. This study confirms the necessity to build an ensemble model using nonlinear machine learning algorithms, topographic descriptors, and separating molecules into charged and neutral compounds to predict E with precision.

Ecotoxicity of pesticide formulations and their mixtures: the case of potato crops in Costa Rica.

Despite their environmental implications, ecotoxicological information regarding pesticide mixtures is relatively scarce. This study aimed to determine the ecotoxicity of individual pesticide formulations and their mixtures (insecticides and fungicides), which are applied during the production cycle of potato, according to agricultural practices from a Latin American region in Costa Rica. Two benchmark organisms were employed: Daphnia magna and Lactuca sativa. First, the evaluation of individual formulations (chlorothalonil, propineb, deltamethrin+imidacloprid, ziram, thiocyclam and chlorpyrifos) revealed differences between available EC50 for active ingredients (a.i.) and their respective formulations toward D. magna; on the contrary, no information could be retrieved from scientific literature for comparison in the case of L. sativa. In general, acute toxicity was higher toward D. magna than L. sativa. Moreover, interactions could not be determined on L. sativa, as the chlorothalonil formulation was not toxic at high levels and the concentration-response to propineb could not be fitted to obtain an IC50 value. The commercial formulation composed of deltamethrin+imidacloprid followed the concentration addition model (when compared with parameters retrieved from individual a.i.) and the other three mixtures evaluated (I: chlorothalonil-propineb-deltamethrin+imidacloprid; II: chlorothalonil-propineb-ziram-thiocyclam; III: chlorothalonil-propineb-chlorpyrifos) produced an antagonistic effect on D. magna, thus suggesting less acute toxicity than their individual components. Subsequent chronic studies showed that one of the most toxic mixtures (II) negatively affected D. magna reproduction at sublethal concentrations indicating that this mixture poses a risk to this species if these pesticides co-exist in freshwater systems. These findings provide useful data to better estimate the impact of real agricultural practices related to the use of agrochemicals.

Elucidating the Racemization Mechanism of Aliphatic and Aromatic Amino Acids by In Silico Tools.

The racemization of biomolecules in the active site can reduce the biological activity of drugs, and the mechanism involved in this process is still not fully comprehended. The present study investigates the impact of aromaticity on racemization using advanced theoretical techniques based on density functional theory. Calculations were performed at the ωb97xd/6-311++g(d,p) level of theory. A compelling explanation for the observed aromatic stabilization via resonance is put forward, involving a carbanion intermediate. The analysis, employing Hammett's parameters, convincingly supports the presence of a negative charge within the transition state of aromatic compounds. Moreover, the combined utilization of natural bond orbital (NBO) analysis and intrinsic reaction coordinate (IRC) calculations confirms the pronounced stabilization of electron distribution within the carbanion intermediate. To enhance our understanding of the racemization process, a thorough examination of the evolution of NBO charges and Wiberg bond indices (WBIs) at all points along the IRC profile is performed. This approach offers valuable insights into the synchronicity parameters governing the racemization reactions.

Hydrogel for the Controlled Delivery of Bioactive Components from Extracts of Eupatorium glutinosum Lam. Leaves.

This research reported a hydrogel loaded with the ethanolic and methanolic extracts of Eupatorium glutinosum Lam. The E. glutinosum extracts were characterized by phytochemical screening, Fourier-transform infrared spectroscopy (FTIR), thin-layer chromatography (TLC), and UV/Vis profile identification. This research also evaluated the pharmacological activity of the extracts using antimicrobial, antioxidant, and anti-inflammatory assays prior to polymeric encapsulation. Results indicate that extracts inhibit the Escherichia colii DH5-α (Gram negative) growth; excellent antioxidant activity was evaluated by the ferric reducing power and total antioxidant activity assays, and extracts showed an anti-hemolytic effect. Moreover, the cotton and microcrystalline cellulose hydrogels demonstrate successful encapsulation based on characterization and kinetics studies such as FTIR, extract release, and swelling degree. Moreover, effective antibacterial activity was registered by the loaded hydrogel. The overall results encourage and show that Eupatorium glutinosum-loaded hydrogel may find a wide range of bandage and wound healing applications in the biomedical area.

On-farm Occurrence of Pharmaceuticals and Their Environmental Hazard: Case Study of a Tropical Dairy farm.

Animal husbandry wastewaters represent an important source of pharmaceuticals into the environment. This work aimed to evaluate the occurrence of pharmaceuticals and their hazard in wastewater from a model dairy farm from Costa Rica. Among the seven pharmaceuticals detected (acetaminophen, caffeine, carbamazepine, ibuprofen, ketoprofen, risperidone, sulfamethazine), caffeine, ibuprofen and acetaminophen showed the highest concentrations, while caffeine, carbamazepine and risperidone were the most frequently detected compounds. High (HQ ≥ 1) or medium (0.1 ≤ HQ < 1) hazard were estimated for three (caffeine, ibuprofen, risperidone) and two (acetaminophen, ketoprofen) pharmaceuticals, respectively; similarly, high overall hazard (∑HQ) and significant ecotoxicity were determined in samples from all sampling points. According to our results, the release of these aqueous matrices is a matter of environmental concern, as the treated wastewater is used for farm irrigation or directly released into nearby water streams. This work contributes to the knowledge on the scarcely described occurrence and risk of pharmaceuticals in Latin American regions.

Na⋯B bond in NaBH 3 - : An induced spin-polarized bond.

The nature of the Na⋯B bond, in the recently synthesized NaBH 3 - adduct, is analyzed on the light of the Na- propensity to polarize along the bond axis as a consequence of the electric field produced by the BH3 fragment. The observed induced polarization has two consequences: (i) the energetic stabilization of the Na- , and (ii) the split of its valence electrons into two opposite lobes along the bond axis. Additionally, an analysis of the electron localization is presented using the information content of the correlated conditional pair density that reveals a significant delocalization between one lobe of the polarized Na- anion and the BH3 fragment at the equilibrium distance. Our findings reported here complement previous works on this system.

Effect of the Nucleophile's Nature on Chloroacetanilide Herbicides Cleavage Reaction Mechanism. A DFT Study.

In this study, the degradation mechanism of chloroacetanilide herbicides in the presence of four different nucleophiles, namely: Br-, I-, HS-, and S2O3-2, was theoretically evaluated using the dispersion-corrected hybrid functional wB97XD and the DGDZVP as a basis set. The comparison of computed activation energies with experimental data shows an excellent correlation (R2 = 0.98 for alachlor and 0.97 for propachlor). The results suggest that the best nucleophiles are those where a sulfur atom performs the nucleophilic attack, whereas the other species are less reactive. Furthermore, it was observed that the different R groups of chloroacetanilide herbicides have a negligible effect on the activation energy of the process. Further insights into the mechanism show that geometrical changes and electronic rearrangements contribute 60% and 40% of the activation energy, respectively. A deeper analysis of the reaction coordinate was conducted, employing the evolution chemical potential, hardness, and electrophilicity index, as well as the electronic flux. The charge analysis shows that the electron density of chlorine increases as the nucleophilic attack occurs. Finally, NBO analysis indicates that the nucleophilic substitution in chloroacetanilides is an asynchronous process with a late transition state for all models except for the case of the iodide attack, which occurs through an early transition state in the reaction.

In Silico Screening of the DrugBank Database to Search for Possible Drugs against SARS-CoV-2.

Coronavirus desease 2019 (COVID-19) is responsible for more than 1.80 M deaths worldwide. A Quantitative Structure-Activity Relationships (QSAR) model is developed based on experimental pIC50 values reported for a structurally diverse dataset. A robust model with only five descriptors is found, with values of R2 = 0.897, Q2LOO = 0.854, and Q2ext = 0.876 and complying with all the parameters established in the validation Tropsha's test. The analysis of the applicability domain (AD) reveals coverage of about 90% for the external test set. Docking and molecular dynamic analysis are performed on the three most relevant biological targets for SARS-CoV-2: main protease, papain-like protease, and RNA-dependent RNA polymerase. A screening of the DrugBank database is executed, predicting the pIC50 value of 6664 drugs, which are IN the AD of the model (coverage = 79%). Fifty-seven possible potent anti-COVID-19 candidates with pIC50 values > 6.6 are identified, and based on a pharmacophore modelling analysis, four compounds of this set can be suggested as potent candidates to be potential inhibitors of SARS-CoV-2. Finally, the biological activity of the compounds was related to the frontier molecular orbitals shapes.

Synthesis, In Vitro, and In Silico Analysis of the Antioxidative Activity of Dapsone Imine Derivatives.

Dapsone (DDS) is an antibacterial drug with well-known antioxidant properties. However, the antioxidant behavior of its derivatives has not been well explored. In the present work, the antioxidant activity of 10 dapsone derivatives 4-substituted was determined by an evaluation in two in vitro models (DPPH radical scavenging assay and ferric reducing antioxidant power). These imine derivatives 1-10 were obtained through condensation between DDS and the corresponding aromatic aldehydes 4-substuited. Three derivatives presented better results than DDS in the determination of DPPH (2, 9, and 10). Likewise, we have three compounds with better reducing activity than dapsone (4, 9, and 10). In order to be more insight, the redox process, a conceptual DFT analysis was carried out. Molecular descriptors such as electronic distribution, the total charge accepting/donating capacity (I/A), and the partial charge accepting/donating capacity (ω+/ω-) were calculated to analyze the relative donor-acceptor capacity through employing a donor acceptor map (DAM). The DFT calculation allowed us to establish a relationship between GAPHOMO-LUMO and DAM with the observed antioxidant effects. According to the results, we concluded that compounds 2 and 3 have the lowest Ra values, representing a good antioxidant behavior observed experimentally in DPPH radical capturing. On the other hand, derivatives 4, 9, and 10 display the best reducing capacity activity with the highest ω- and Rd values. Consequently, we propose these compounds as the best antireductants in our DDS imine derivative series.

Ensemble Models Based on QuBiLS-MAS Features and Shallow Learning for the Prediction of Drug-Induced Liver Toxicity: Improving Deep Learning and Traditional Approaches.

Drug-induced liver injury (DILI) is a key safety issue in the drug discovery pipeline and a regulatory concern. Thus, many in silico tools have been proposed to improve the hepatotoxicity prediction of organic-type chemicals. Here, classifiers for the prediction of DILI were developed by using QuBiLS-MAS 0-2.5D molecular descriptors and shallow machine learning techniques, on a training set composed of 1075 molecules. The best ensemble model build, E13, was obtained with good statistical parameters for the learning series, namely, the following: accuracy = 0.840, sensibility = 0.890, specificity = 0.761, Matthew's correlation coefficient = 0.660, and area under the ROC curve = 0.904. The model was also satisfactorily evaluated with Y-scrambling test, and repeated k-fold cross-validation and repeated k-holdout validation. In addition, an exhaustive external validation was also carried out by using two test sets and five external test sets, with an average accuracy value equal to 0.854 (±0.062) and a coverage equal to 98.4% according to its applicability domain. A statistical comparison of the performance of the E13 model, with regard to results and tools (e.g., Padel DDPredictor Software, Deep Learning DILIserver, and Vslead) reported in the literature, was also performed. In general, E13 presented the best global performance in all experiments. The sum of the ranking differences procedure provided a very similar grouping pattern to that of the M-ANOVA statistical analysis, where E13 was identified as the best model for DILI predictions. A noncommercial and fully cross-platform software for the DILI prediction was also developed, which is freely available at http://tomocomd.com/apps/ptoxra. This software was used for the screening of seven data sets, containing natural products, leads, toxic materials, and FDA approved drugs, to assess the usefulness of the QSAR models in the DILI labeling of organic substances; it was found that 50-92% of the evaluated molecules are positive-DILI compounds. All in all, it can be stated that the E13 model is a relevant method for the prediction of DILI risk in humans, as it shows the best results among all of the methods analyzed.

A case of intraplacental gestational choriocarcinoma; characterised by the methylation pattern of the early placenta and an absence of driver mutations.

BACKGROUND: Gestational choriocarcinoma is a rare malignancy believed to arise from the trophoblast cells of the placenta. Despite the frequently aggressive clinical nature, choriocarcinoma has been routinely curable with cytotoxic chemotherapy for over 50 years. To date little is known regarding the route to oncogenesis in this malignancy. METHODS: In a case of intraplacental choriocarcinoma, we have performed detailed genetic studies including microsatellite analysis, whole genome sequencing (WGS) and methylation analysis of the tumour and surrounding mature placenta. RESULTS: The results of the WGS sequencing indicated a very low level of mutation and the absence of any driver mutations or oncogene activity in the tumour. The methylation analysis identified a distinctly different profile in the tumour from that of the mature placenta. Comparison with a panel of reference methylation profiles from different stages of placental development indicated that the tumour segregated with the first trimester samples. CONCLUSIONS: These findings suggest that gestational choriocarcinoma is likely to arise as a result of aberrations of methylation during development, rather than from DNA mutations. The results support the hypothesis that gestational choriocarcinoma arises from a normally transient early trophoblast cell. At this point in development this cell naturally has a phenotype of rapid division, tissue invasion and sensitivity to DNA damaging chemotherapy that is very similar to that of the mature choriocarcinoma cell.

Modeling the Antileukemia Activity of Ellipticine-Related Compounds: QSAR and Molecular Docking Study.

The antileukemia cancer activity of organic compounds analogous to ellipticine representes a critical endpoint in the understanding of this dramatic disease. A molecular modeling simulation on a dataset of 23 compounds, all of which comply with Lipinski's rules and have a structure analogous to ellipticine, was performed using the quantitative structure activity relationship (QSAR) technique, followed by a detailed docking study on three different proteins significantly involved in this disease (PDB IDs: SYK, PI3K and BTK). As a result, a model with only four descriptors (HOMO, softness, AC1RABAMBID, and TS1KFABMID) was found to be robust enough for prediction of the antileukemia activity of the compounds studied in this work, with an R2 of 0.899 and Q2 of 0.730. A favorable interaction between the compounds and their target proteins was found in all cases; in particular, compounds 9 and 22 showed high activity and binding free energy values of around -10 kcal/mol. Theses compounds were evaluated in detail based on their molecular structure, and some modifications are suggested herein to enhance their biological activity. In particular, compounds 22_1, 22_2, 9_1, and 9_2 are indicated as possible new, potent ellipticine derivatives to be synthesized and biologically tested.

New Insight into the Chloroacetanilide Herbicide Degradation Mechanism through a Nucleophilic Attack of Hydrogen Sulfide.

The nucleophilic attack of hydrogen sulfide (HS-) on six different chloroacetanilide herbicides was evaluated theoretically using the dispersion-corrected hybrid functional wB97XD and the 6-311++G(2d,2p) Pople basis sets. The six evaluated substrates were propachlor (A), alachlor (B), metolachlor (C), tioacetanilide (D), ß-anilide (E), and methylene (F). Three possible mechanisms were considered: (a) bimolecular nucleophilic substitution (SN²) reaction mechanism, (b) oxygen assistance, and (c) nitrogen assistance. Mechanisms based on O- and N-assistance were discarded due to a very high activation barrier in comparison with the corresponding SN² mechanism, with the exception of compound F. The N-assistance mechanism for compound F had a free activation energy of 23.52 kcal/mol, which was close to the value for the corresponding SN² mechanism (23.94 kcal/mol), as these two mechanisms could occur in parallel reactions with almost 50% of each one. In compounds A to D, an important electron-withdrawing effect of the C=O and C=S groups was seen, and consequently, the activation free energies in these SN² reactions were smaller, with a value of approximately 18 kcal/mol. Instead, compounds E and F, which have a CH2 group in the ß-position, presented a higher activation free energy (≈22 kcal/mol). Good agreement was found between experimental and theoretical values for all cases, and a reaction force analysis was performed on the intrinsic reaction coordinate profile in order to gain more details about the reaction mechanism. Finally, from the natural bond orbital (NBO) analysis, it was possible to evaluate the electronic reorganization through the reaction pathway where all the transition states were early in nature in the reaction coordinate (δBav < 50%); the transition states corresponding to compounds A to D turned out to be more synchronous than those for compounds E and F.

Classical QSAR and Docking Simulation of 4-Pyridone Derivatives for Their Antimalarial Activity.

In this work, the minimum energy structures of 22 4-pyridone derivatives have been optimized at Density Functional Theory level, and several quantum molecular, including electronic and thermodynamic descriptors, were computed for these substrates in order to obtain a statistical and meaningful QSAR equation. In this sense, by using multiple linear regressions, five mathematical models have been obtained. The best model with only four descriptors (r² = 0.86, Q² = 0.92, S.E.P = 0.38) was validated by the leave-one-out cross-validation method. The antimalarial activity can be explained by the combination of the four mentioned descriptors e.g., electronic potential, dipolar momentum, partition coefficient and molar refractivity. The statistical parameters of this model suggest that it is robust enough to predict the antimalarial activity of new possible compounds; consequently, three small chemical modifications into the structural core of these compounds were performed specifically on the most active compound of the series (compound 13). These three new suggested compounds were leveled as 13A, 13B and 13C, and the predicted biological antimalarial activity is 0.02 µM, 0.03 µM, and 0.07 µM, respectively. In order to complement these results focused on the possible action mechanism of the substrates, a docking simulation was included for these new structures as well as for the compound 13 and the docking scores (binding affinity) obtained for the interaction of these substrates with the cytochrome bc1, were -7.5, -7.2, -6.9 and -7.5 kcal/mol for 13A, 13B, 13C and compound 13, respectively, which suggests that these compounds are good candidates for its biological application in this illness.

Integrated QTL detection for key breeding traits in multiple peach progenies.

BACKGROUND: Peach (Prunus persica (L.) Batsch) is a major temperate fruit crop with an intense breeding activity. Breeding is facilitated by knowledge of the inheritance of the key traits that are often of a quantitative nature. QTLs have traditionally been studied using the phenotype of a single progeny (usually a full-sib progeny) and the correlation with a set of markers covering its genome. This approach has allowed the identification of various genes and QTLs but is limited by the small numbers of individuals used and by the narrow transect of the variability analyzed. In this article we propose the use of a multi-progeny mapping strategy that used pedigree information and Bayesian approaches that supports a more precise and complete survey of the available genetic variability. RESULTS: Seven key agronomic characters (data from 1 to 3 years) were analyzed in 18 progenies from crosses between occidental commercial genotypes and various exotic lines including accessions of other Prunus species. A total of 1467 plants from these progenies were genotyped with a 9 k SNP array. Forty-seven QTLs were identified, 22 coinciding with major genes and QTLs that have been consistently found in the same populations when studied individually and 25 were new. A substantial part of the QTLs observed (47%) would not have been detected in crosses between only commercial materials, showing the high value of exotic lines as a source of novel alleles for the commercial gene pool. Our strategy also provided estimations on the narrow sense heritability of each character, and the estimation of the QTL genotypes of each parent for the different QTLs and their breeding value. CONCLUSIONS: The integrated strategy used provides a broader and more accurate picture of the variability available for peach breeding with the identification of many new QTLs, information on the sources of the alleles of interest and the breeding values of the potential donors of such valuable alleles. These results are first-hand information for breeders and a step forward towards the implementation of DNA-informed strategies to facilitate selection of new cultivars with improved productivity and quality.

Transforming a Stable Amide into a Highly Reactive One: Capturing the Essence of Enzymatic Catalysis.

Aspartic proteinases, which include HIV-1 proteinase, function with two aspartate carboxy groups at the active site. This relationship has been modeled in a system possessing an otherwise unactivated amide positioned between two carboxy groups. The model amide is cleaved at an enzyme-like rate that renders the amide nonisolable at 35 °C and pH 4 owing to the joint presence of carboxy and carboxylate groups. A currently advanced theory attributing almost the entire catalytic power of enzymes to electrostatic reorganization is shown to be superfluous when suitable interatomic interactions are present. Our kinetic results are consistent with spatiotemporal concepts where embedding the amide group between two carboxylic moieties in proper geometries, at distances less than the diameter of water, leads to enzyme-like rate enhancements. Space and time are the essence of enzyme catalysis.

New light on phosphate transfer from triesters.

The reactivity of triesters is discussed in the general context of phosphate transfer, as usually studied for the reactions of mono- and diesters. Systematic work has typically concentrated on the Linear Free Energy Relationships measuring the dependence of reactivity on the nucleophile and the leaving group, but new results indicate that it can depend equally strongly on the two non-leaving (sometimes known as spectator) groups. This conclusion is supported by first results from theoretical calculations: which also predict that a two-step mechanism can be favored over a concerted S(N)2(P) mechanism even for reactions involving leaving groups as good as p-nitrophenolate. This article is part of a Special Issue entitled: Chemistry and mechanism of phosphatases, diesterases and triesterases.

Unraveling the Hemolytic Toxicity Tapestry of Peptides using Chemical Space Complex Networks.

Peptides have emerged as promising therapeutic agents. However, their potential is hindered by hemotoxicity. Understanding the hemotoxicity of peptides is crucial for developing safe and effective peptide-based therapeutics. Here, we employed chemical space complex networks (CSNs) to unravel the hemotoxicity tapestry of peptides. CSNs are powerful tools for visualizing and analyzing the relationships between peptides based on their physicochemical properties and structural features. We constructed CSNs from the StarPepDB database, encompassing 2004 hemolytic peptides, and explored the impact of seven different (dis)similarity measures on network topology and cluster (communities) distribution. Our findings revealed that each CSN extracts orthogonal information, enhancing the motif discovery and enrichment process. We identified 12 consensus hemolytic motifs, whose amino acid composition unveiled a high abundance of lysine, leucine, and valine residues, while aspartic acid, methionine, histidine, asparagine and glutamine were depleted. Additionally, physicochemical properties were used to characterize clusters/communities of hemolytic peptides. To predict hemolytic activity directly from peptide sequences, we constructed multi-query similarity searching models (MQSSMs), which outperformed cutting-edge machine learning (ML)-based models, demonstrating robust hemotoxicity prediction capabilities. Overall, this novel in silico approach uses complex network science as its central strategy to develop robust model classifiers, to characterize the chemical space and to discover new motifs from hemolytic peptides. This will help to enhance the design/selection of peptides with potential therapeutic activity and low toxicity.