Prediction of HIV sensitivity to monoclonal antibodies using aminoacid sequences and deep learning.

MOTIVATION: Knowing the sensitivity of a viral strain versus a monoclonal antibody is of interest for HIV vaccine development and therapy. The HIV strains vary in their resistance to antibodies, and the accurate prediction of virus-antibody sensitivity can be used to find potent antibody combinations that broadly neutralize multiple and diverse HIV strains. Sensitivity prediction can be combined with other methods such as generative algorithms to design novel antibodies in silico or with feature selection to uncover the sites of interest in the sequence. However, these tools are limited in the absence of in silico accurate prediction methods. RESULTS: Our method leverages the CATNAP dataset, probably the most comprehensive collection of HIV-antibodies assays, and predicts the antibody-virus sensitivity in the form of binary classification. The methods proposed by others focus primarily on analyzing the virus sequences. However, our article demonstrates the advantages gained by modeling the antibody-virus sensitivity as a function of both virus and antibody sequences. The input is formed by the virus envelope and the antibody variable region aminoacid sequences. No structural features are required, which makes our system very practical, given that sequence data is more common than structures. We compare with two other state-of-the-art methods that leverage the same dataset and use sequence data only. Our approach, based on neuronal networks and transfer learning, measures increased predictive performance as measured on a set of 31 specific broadly neutralizing antibodies. AVAILABILITY AND IMPLEMENTATION: https://github.com/vlad-danaila/deep_hiv_ab_pred/tree/fc-att-fix.

Synthesis and Bioinformatic Characterization of New Schiff Bases with Possible Applicability in Brain Disorders.

(1) Background: The research aims to find new treatments for neurodegenerative diseases, in particular, Alzheimer's disease. (2) Methods: This article presents a bioinformatics and pathology study of new Schiff bases, (EZ)-N'-benzylidene-(2RS)-2-(6-chloro-9H-carbazol-2-yl)propanehydrazide derivatives, and aims to evaluate the drug-like, pharmacokinetic, pharmacodynamic and pharmacogenomic properties, as well as to predict the binding to therapeutic targets by applying bioinformatics, cheminformatics and computational pharmacological methods. (3) Results: We obtained these Schiff bases by condensing (2RS)-2-(6-chloro-9H-carbazol-2-yl)propanehydrazide with aromatic aldehydes, using the advantages of microwave irradiation. The newly synthesized compounds were characterized spectrally, using FT-IR and NMR spectroscopy, which confirmed their structure. Using bioinformatics tools, we noticed that all new compounds are drug-likeness features and may be proposed as potentially neuropsychiatric drugs (4) Conclusions: Using bioinformatics tools, we determined that the new compound 1e had a high potential to be used as a good candidate in neurodegenerative disorders treatment.

Learning the Relationship between the Primary Structure of HIV Envelope Glycoproteins and Neutralization Activity of Particular Antibodies by Using Artificial Neural Networks.

The dependency between the primary structure of HIV envelope glycoproteins (ENV) and the neutralization data for given antibodies is very complicated and depends on a large number of factors, such as the binding affinity of a given antibody for a given ENV protein, and the intrinsic infection kinetics of the viral strain. This paper presents a first approach to learning these dependencies using an artificial feedforward neural network which is trained to learn from experimental data. The results presented here demonstrate that the trained neural network is able to generalize on new viral strains and to predict reliable values of neutralizing activities of given antibodies against HIV-1.

Structure-biological function relationship extended to mitotic arrest-deficient 2-like protein Mad2 native and mutants-new opportunity for genetic disorder control.

Overexpression of mitotic arrest-deficient proteins Mad1 and Mad2, two components of spindle assembly checkpoint, is a risk factor for chromosomal instability (CIN) and a trigger of many genetic disorders. Mad2 transition from inactive open (O-Mad2) to active closed (C-Mad2) conformations or Mad2 binding to specific partners (cell-division cycle protein 20 (Cdc20) or Mad1) were targets of previous pharmacogenomics studies. Here, Mad2 binding to Cdc20 and the interconversion rate from open to closed Mad2 were predicted and the molecular features with a critical contribution to these processes were determined by extending the quantitative structure-activity relationship (QSAR) method to large-size proteins such as Mad2. QSAR models were built based on available published data on 23 Mad2 mutants inducing CIN-related functional changes. The most relevant descriptors identified for predicting Mad2 native and mutants action mechanism and their involvement in genetic disorders are the steric (van der Waals area and solvent accessible area and their subdivided) and energetic van der Waals energy descriptors. The reliability of our QSAR models is indicated by significant values of statistical coefficients: Cross-validated correlation q2 (0.53-0.65) and fitted correlation r2 (0.82-0.90). Moreover, based on established QSAR equations, we rationally design and analyze nine de novo Mad2 mutants as possible promoters of CIN.

Therapeutic Potentials of Aconite-like Alkaloids: Bioinformatics and Experimental Approaches.

Compounds from plants that are used in traditional medicine may have medicinal properties. It is well known that plants belonging to the genus Aconitum are highly poisonous. Utilizing substances derived from Aconitum sp. has been linked to negative effects. In addition to their toxicity, the natural substances derived from Aconitum species may have a range of biological effects on humans, such as analgesic, anti-inflammatory, and anti-cancer characteristics. Multiple in silico, in vitro, and in vivo studies have demonstrated the effectiveness of their therapeutic effects. In this review, the clinical effects of natural compounds extracted from Aconitum sp., focusing on aconitelike alkaloids, are investigated particularly by bioinformatics tools, such as the quantitative structure- activity relationship method, molecular docking, and predicted pharmacokinetic and pharmacodynamic profiles. The experimental and bioinformatics aspects of aconitine's pharmacogenomic profile are discussed. Our review could help shed light on the molecular mechanisms of Aconitum sp. compounds. The effects of several aconite-like alkaloids, such as aconitine, methyllycacintine, or hypaconitine, on specific molecular targets, including voltage-gated sodium channels, CAMK2A and CAMK2G during anesthesia, or BCL2, BCL-XP, and PARP-1 receptors during cancer therapy, are evaluated. According to the reviewed literature, aconite and aconite derivatives have a high affinity for the PARP-1 receptor. The toxicity estimations for aconitine indicate hepatotoxicity and hERG II inhibitor activity; however, this compound is not predicted to be AMES toxic or an hERG I inhibitor. The efficacy of aconitine and its derivatives in treating many illnesses has been proven experimentally. Toxicity occurs as a result of the high ingested dose; however, the usage of this drug in future research is based on the small quantity of an active compound that fulfills a therapeutic role.

Repurposing anti-inflammatory drugs for fighting planktonic and biofilm growth. New carbazole derivatives based on the NSAID carprofen: synthesis, in silico and in vitro bioevaluation.

Introduction: One of the promising leads for the rapid discovery of alternative antimicrobial agents is to repurpose other drugs, such as nonsteroidal anti-inflammatory agents (NSAIDs) for fighting bacterial infections and antimicrobial resistance. Methods: A series of new carbazole derivatives based on the readily available anti-inflammatory drug carprofen has been obtained by nitration, halogenation and N-alkylation of carprofen and its esters. The structures of these carbazole compounds were assigned by NMR and IR spectroscopy. Regioselective electrophilic substitution by nitration and halogenation at the carbazole ring was assigned from H NMR spectra. The single crystal X-ray structures of two representative derivatives obtained by dibromination of carprofen, were also determined. The total antioxidant capacity (TAC) was measured using the DPPH method. The antimicrobial activity assay was performed using quantitative methods, allowing establishment of the minimal inhibitory/bactericidal/biofilm eradication concentrations (MIC/MBC/MBEC) on Gram-positive (Staphylococcus aureus, Enterococcus faecalis) and Gram-negative (Escherichia coli, Pseudomonas aeruginosa) strains. Computational assays have been performed to assess the drug- and lead-likeness, pharmacokinetics (ADME-Tox) and pharmacogenomics profiles. Results and discussion: The crystal X-ray structures of 3,8-dibromocarprofen and its methyl ester have revealed significant differences in their supramolecular assemblies. The most active antioxidant compound was 1i, bearing one chlorine and two bromine atoms, as well as the CO2Me group. Among the tested derivatives, 1h bearing one chlorine and two bromine atoms has exhibited the widest antibacterial spectrum and the most intensive inhibitory activity, especially against the Gram-positive strains, in planktonic and biofilm growth state. The compounds 1a (bearing one chlorine, one NO2 and one CO2Me group) and 1i (bearing one chlorine, two bromine atoms and a CO2Me group) exhibited the best antibiofilm activity in the case of the P. aeruginosa strain. Moreover, these compounds comply with the drug-likeness rules, have good oral bioavailability and are not carcinogenic or mutagenic. The results demonstrate that these new carbazole derivatives have a molecular profile which deserves to be explored further for the development of novel antibacterial and antibiofilm agents.

The applications of machine learning in HIV neutralizing antibodies research-A systematic review.

Machine learning algorithms play an essential role in bioinformatics and allow exploring the vast and noisy biological data in unrivaled ways. This paper is a systematic review of the applications of machine learning in the study of HIV neutralizing antibodies. This significant and vast research domain can pave the way to novel treatments and to a vaccine. We selected the relevant papers by investigating the available literature from the Web of Science and PubMed databases in the last decade. The computational methods are applied in neutralization potency prediction, neutralization span prediction against multiple viral strains, antibody-virus binding sites detection, enhanced antibodies design, and the study of the antibody-induced immune response. These methods are viewed from multiple angles spanning data processing, model description, feature selection, evaluation, and sometimes paper comparisons. The algorithms are diverse and include supervised, unsupervised, and generative types. Both classical machine learning and modern deep learning were taken into account. The review ends with our ideas regarding future research directions and challenges.

3D-ALMOND-QSAR Models to Predict the Antidepressant Effect of Some Natural Compounds.

The current treatment of depression involves antidepressant synthetic drugs that have a variety of side effects. In searching for alternatives, natural compounds could represent a solution, as many studies reported that such compounds modulate the nervous system and exhibit antidepressant effects. We used bioinformatics methods to predict the antidepressant effect of ten natural compounds with neuroleptic activity, reported in the literature. For all compounds we computed their drug-likeness, absorption, distribution, metabolism, excretion (ADME), and toxicity profiles. Their antidepressant and neuroleptic activities were predicted by 3D-ALMOND-QSAR models built by considering three important targets, namely serotonin transporter (SERT), 5-hydroxytryptamine receptor 1A (5-HT1A), and dopamine D2 receptor. For our QSAR models we have used the following molecular descriptors: hydrophobicity, electrostatic, and hydrogen bond donor/acceptor. Our results showed that all compounds present drug-likeness features as well as promising ADME features and no toxicity. Most compounds appear to modulate SERT, and fewer appear as ligands for 5-HT1A and D2 receptors. From our prediction, linalyl acetate appears as the only ligand for all three targets, neryl acetate appears as a ligand for SERT and D2 receptors, while 1,8-cineole appears as a ligand for 5-HT1A and D2 receptors.

Chondroitin sulfate proteoglycans: structure-function relationship with implication in neural development and brain disorders.

Chondroitin sulfate proteoglycans (CSPGS) are extracellular matrix components that contain two structural parts with distinct functions: a protein core and glycosaminoglycan (GAG) side chains. CSPGs are known to be involved in important cell processes like cell adhesion and growth, receptor binding, or cell migration. It is recognized that the presence of CSPGs is critical in neuronal growth mechanisms including axon guidance following injury of nervous system components such as spinal cord and brain. CSPGs are upregulated in the central nervous system after injury and participate in the inhibition of axon regeneration mainly through their GAG side chains. Recently, it was shown that some CSPGs members like aggrecan, versican, and neurocan were strongly involved in brain disorders like bipolar disorder (BD), schizophrenia, and ADHD. In this paper, we present the chemical structure-biological functions relationship of CSPGs, both in health state and in genetic disorders, addressing methods represented by genome-wide and crystallographic data as well as molecular modeling and quantitative structure-activity relationship.

More effective DPP4 inhibitors as antidiabetics based on sitagliptin applied QSAR and clinical methods.

Xanthine-based molecules such as serine protease dipeptidyl peptidase 4 (DPP4) inhibitors are compounds often used in improving glycemic control in type 2 diabetic patients and also used for their effects as mild stimulants and as bronchodilators, notably in treating asthma symptoms. Here, we aim to better understand the molecular features affecting activity of xanthine-based DPP4 inhibitors such as sitagliptin and related compounds and use these features to de novo predict improved sitagliptin derivatives. To this end, we performed a clinical study to examine the efficacy and safety of once-daily 100 mg oral sitagliptin as monotherapy in Romanian patients with type 2 diabetes. This study indicates that sitagliptin effectively decreases the glycemic level and provides very good glycemic equilibrium. To predict putative new drugs with identical pharmacological effects at lower dosages, we generate QSAR models based on compound series containing 35 DPP4 inhibitors. We establish that the physicochemical parameters critical for DPP4 inhibitory activity are: hydrophobicity described by the logarithm of the octanol/water partition coefficient, counts of rotatable bonds, hydrogen bond donor and acceptor atoms, and topological polar surface area. The predictive power of our QSAR models is indicated by significant values of statistical coefficients: cross-validated correlation q2 (0.77), fitted correlation coefficient r2 (0.85) and standard error of prediction (0.34). Based on the established QSAR equations, we propose and analyse 19 new sitagliptin derivatives with possibly improved pharmacological effect as DPP4 inhibitors.

More effective antimicrobial mastoparan derivatives, generated by 3D-QSAR-Almond and computational mutagenesis.

Antimicrobial peptides are drugs used against a wide range of pathogens which present a great advantage: in contrast with antibiotics they do not develop resistance. The wide spectrum of antimicrobial peptides advertises them in the research and pharmaceutical industry as attractive starting points for obtaining new, more effective analogs. Here we predict the antimicrobial activity against Bacillus subtilis (expressed as minimal inhibitory concentration values) for 33 mastoparan analogs and their new derivatives by a non-aligned 3D-QSAR (quantitative structure-activity relationship) method. We establish the contribution to antimicrobial activity of molecular descriptors (hydrophobicity, hydrogen bond donor and steric), correlated with contributions from the membrane environment (sodium, potassium, chloride). Our best QSAR models show significant cross-validated correlation q(2) (0.55-0.75), fitted correlation r(2) (greater than 0.90) coefficients and standard error of prediction SDEP (less than 0.250). Moreover, based on our most accurate 3D-QSAR models, we propose nine new mastoparan analogs, obtained by computational mutagenesis, some of them predicted to have significantly improved antimicrobial activity compared to the parent compound.

A software tool for modeling and simulation of numerical P systems.

UNLABELLED: A P system represents a distributed and parallel bio-inspired computing model in which basic data structures are multi-sets or strings. Numerical P systems have been recently introduced and they use numerical variables and local programs (or evolution rules), usually in a deterministic way. They may find interesting applications in areas such as computational biology, process control or robotics. The first simulator of numerical P systems (SNUPS) has been designed, implemented and made available to the scientific community by the authors of this paper. SNUPS allows a wide range of applications, from modeling and simulation of ordinary differential equations, to the use of membrane systems as computational blocks of cognitive architectures, and as controllers for autonomous mobile robots. This paper describes the functioning of a numerical P system and presents an overview of SNUPS capabilities together with an illustrative example. AVAILABILITY: SNUPS is freely available to researchers as a standalone application and may be downloaded from a dedicated website, http://snups.ics.pub.ro/, which includes an user manual and sample membrane structures.

3D-QSAR design of new escitalopram derivatives for the treatment of major depressive disorders.

Antidepressants are psychiatric agents used for the treatment of different types of depression being at present amongst the most commonly prescribed drug, while their effectiveness and adverse effects are the subject of many studies and competing claims. Having studied five QSAR models predicting the biological activities of 18 antidepressants, already approved for clinical treatment, in interaction with the serotonin transporter (SERT), we attempted to establish the membrane ionsâ contributions (sodium, potassium, chlorine and calcium) supplied by donor/acceptor hydrogen bond character and electrostatic field to the antidepressant activity. Significant cross-validated correlation q(2) (0.5â0.6) and the fitted correlation r(2) (0.7â0.82) coefficients were obtained indicating that the models can predict the antidepressant activity of compounds. Moreover, considering the contribution of membrane ions (sodium, potassium and calcium) and hydrogen bond donor character, we have proposed a library of 24 new escitalopram structures, some of them probably with significantly improved antidepressant activity in comparison with the parent compound.