Search

Paediatric Acquired ImmunoDeficiency Syndrome (AIDS) is a life-threatening and infectious disease in which the Human Immunodeficiency Virus (HIV) is mainly transmitted through Mother-To- Child Transmission (MTCT) during pregnancy, labour and delivery, or breastfeeding. This review provides an overview of the distinct therapeutic alternatives to abolish the systemic viral replication in paediatric HIV-1 infection. Numerous classes of antiretroviral agents have emerged as therapeutic tools for downregulation of different steps in the HIV replication process. These classes encompass Non- Nucleoside Analogue Reverse Transcriptase Inhibitors (NNRTIs), Nucleoside/Nucleotide Analogue Reverse Transcriptase Inhibitors (NRTIs/NtRTIs), INtegrase Inhibitors (INIs), Protease Inhibitors (PIs), and Entry Inhibitors (EIs). Co-administration of certain antiretroviral drugs with Pharmacokinetic Enhancers (PEs) may boost the effectiveness of the primary therapeutic agent. The combination of multiple antiretroviral drug regimens (Highly Active AntiRetroviral Therapy - HAART) is currently the standard therapeutic approach for HIV infection. So far, the use of HAART offers the best opportunity for prolonged and maximal viral suppression, and preservation of the immune system upon HIV infection. Still, the frequent administration of high doses of multiple drugs, their inefficient ability to reach the viral reservoirs in adequate doses, the development of drug resistance, and the lack of patient compliance compromise the complete HIV elimination. The development of nanotechnology-based drug delivery systems may enable targeted delivery of antiretroviral agents to inaccessible viral reservoir sites at therapeutic concentrations. In addition, the application of Computer-Aided Drug Design (CADD) approaches has provided valuable tools for the development of anti-HIV drug candidates with favourable pharmacodynamics and pharmacokinetic properties.

High-throughput screening technologies continues to produce large amounts of multiomics data from different populations and cell types for various diseases, such as cancer. However, analysis of such data encounters difficulties due to cancer heterogeneity, further exacerbated by human biological complexity and genomic variability. There is a need to redefine the drug discovery development pipeline, bringing an Artificial Intelligence (AI)-powered informational view that integrates relevant biological information and explores new ways to develop effective anticancer approaches. Here, we show SynPred, an interdisciplinary approach that leverages specifically designed ensembles of AI-algorithms, links omics and biophysical traits to predict synergistic anticancer drug synergy. SynPred exhibits state-of-the-art performance metrics: accuracy – 0.85, precision – 0.77, recall – 0.75, AUROC – 0.82, and F1-score - 0.76 in an independent test set. Moreover, data interpretability was achieved by deploying the most current and robust feature importance approaches. A simple web-based application available online at http://www.moreiralab.com/resources/synpred/ was constructed to predict synergistic anticancer drug combinations requiring only the upload of the two drug SMILES to be tested, allowing easy access by non-expert researchers.

Protein Hot-Spots (HS) are experimentally determined amino acids, key to small ligand binding and tend to be structural landmarks on protein&ndash, protein interactions. As such, they were extensively approached by structure-based Machine Learning (ML) prediction methods. However, the availability of a much larger array of protein sequences in comparison to determined tree-dimensional structures indicates that a sequence-based HS predictor has the potential to be more useful for the scientific community. Herein, we present SPOTONE, a new ML predictor able to accurately classify protein HS via sequence-only features. This algorithm shows accuracy, AUROC, precision, recall and F1-score of 0.82, 0.83, 0.91, 0.82 and 0.85, respectively, on an independent testing set. The algorithm is deployed within a free-to-use webserver at http://moreiralab.com/resources/spotone, only requiring the user to submit a FASTA file with one or more protein sequences.

Targeting protein-protein interactions is a challenge and crucial task of the drug discovery process. A good starting point for rational drug design is the identification of hot spots (HS) at protein-protein interfaces, typically conserved residues that contribute most significantly to the binding. In this chapter, we depict point-by-point an in-house pipeline used for HS prediction using only sequence-based features from the well-known SpotOn dataset of soluble proteins (Moreira et al., Sci Rep 7:8007, 2017), through the implementation of a deep neural network. The presented pipeline is divided into three steps: (1) feature extraction, (2) deep learning classification, and (3) model evaluation. We present all the available resources, including code snippets, the main dataset, and the free and open-source modules/packages necessary for full replication of the protocol. The users should be able to develop an HS prediction model with accuracy, precision, recall, and AUROC of 0.96, 0.93, 0.91, and 0.86, respectively.

Targeting protein-protein interactions is a challenge and crucial task of the drug discovery process. A good starting point for rational drug design is the identification of hot spots (HS) at protein-protein interfaces, typically conserved residues that contribute most significantly to the binding. In this chapter, we depict point-by-point an in-house pipeline used for HS prediction using only sequence-based features from the well-known SpotOn dataset of soluble proteins (Moreira et al., Sci Rep 7:8007, 2017), through the implementation of a deep neural network. The presented pipeline is divided into three steps: (1) feature extraction, (2) deep learning classification, and (3) model evaluation. We present all the available resources, including code snippets, the main dataset, and the free and open-source modules/packages necessary for full replication of the protocol. The users should be able to develop an HS prediction model with accuracy, precision, recall, and AUROC of 0.96, 0.93, 0.91, and 0.86, respectively.

G-Protein coupled receptors (GPCRs) are involved in a myriad of pathways key for human physiology through the formation of complexes with intracellular partners such as G-proteins and arrestins (Arrs). However, the structural and dynamical determinants of these complexes are still largely unknown. Herein, we developed a computational big-data pipeline that enables the structural characterization of GPCR complexes with no available structure. This pipeline was used to study a well-known group of catecholamine receptors, the human dopamine receptor (DXR) family and its complexes, producing novel insights into the physiological properties of these important drug targets. A detailed description of the protein interfaces of all members of the DXR family (D1R, D2R, D3R, D4R, and D5R) and the corresponding protein interfaces of their binding partners (Arrs: Arr2 and Arr3; G-proteins: Gi1, Gi2, Gi3, Go, Gob, Gq, Gslo, Gssh, Gt2, and Gz) was generated. To produce reliable structures of the DXR family in complex with either G-proteins or Arrs, we performed homology modeling using as templates the structures of the β2-adrenergic receptor (β2AR) bound to Gs, the rhodopsin bound to Gi, and the recently acquired neurotensin receptor-1 (NTSR1) and muscarinic 2 receptor (M2R) bound to arrestin (Arr). Among others, the work demonstrated that the three partner groups, Arrs and Gs- and Gi-proteins, are all structurally and dynamically distinct. Additionally, it was revealed the involvement of different structural motifs in G-protein selective coupling between D1- and D2-like receptors. Having constructed and analyzed 50 models involving DXR, this work represents an unprecedented large-scale analysis of GPCR-intracellular partner interface determinants. All data is available at www.moreiralab.com/resources/dxr.

Influenza (flu) is a contagious viral disease, which targets the human respiratory tract and spreads throughout the world each year. Every year, influenza infects around 10% of the world population and between 290,000 and 650,000 people die from it according to the World Health Organization (WHO). Influenza viruses belong to the Orthomyxoviridae family and have a negative sense eight-segment single-stranded RNA genome that encodes 11 different proteins. The only control over influenza seasonal epidemic outbreaks around the world are vaccines, annually updated according to viral strains in circulation, but, because of high rates of mutation and recurrent genetic assortment, new viral strains of influenza are constantly emerging, increasing the likelihood of pandemics. Vaccination effectiveness is limited, calling for new preventive and therapeutic approaches and a better understanding of the virus–host interactions. In particular, grasping the role of influenza non-structural protein 1 (NS1) and related known interactions in the host cell is pivotal to better understand the mechanisms of virus infection and replication, and thus propose more effective antiviral approaches. In this review, we assess the structure of NS1, its dynamics, and multiple functions and interactions, to highlight the central role of this protein in viral biology and its potential use as an effective therapeutic target to tackle seasonal and pandemic influenza.

GPCR oligomerization has emerged as a hot topic in the GPCR field in the last years. Receptors that are part of these oligomers can influence each other's function, although it is not yet entirely understood how these interactions work. The existence of such a highly complex network of interactions between GPCRs generates the possibility of alternative targets for new therapeutic approaches. However, challenges still exist in the characterization of these complexes, especially at the interface level. Different experimental approaches, such as FRET or BRET, are usually combined to study GPCR oligomer interactions. Computational methods have been applied as a useful tool for retrieving information from GPCR sequences and the few X-ray-resolved oligomeric structures that are accessible, as well as for predicting new and trustworthy GPCR oligomeric interfaces. Machine-learning (ML) approaches have recently helped with some hindrances of other methods. By joining and evaluating multiple structure-, sequence- and co-evolution-based features on the same algorithm, it is possible to dilute the issues of particular structures and residues that arise from the experimental methodology into all-encompassing algorithms capable of accurately predict GPCR-GPCR interfaces. All these methods used as a single or a combined approach provide useful information about GPCR oligomerization and its role in GPCR function and dynamics. Altogether, we present experimental, computational and machine-learning methods used to study oligomers interfaces, as well as strategies that have been used to target these dynamic complexes.

Membrane proteins are essential vessels for cell communication both with other cells and noncellular structures. They modulate environment responses and mediate a myriad of biological processes. Dimerization and multimerization processes have been shown to further increase the already high specificity of these processes. Due to their central role in various cell and organism functions, these multimers are often associated with health conditions, such as Alzheimer's disease (AD), Parkinson's disease (PD), and diabetes, among others.Understanding the membrane protein dimers' interface takes advantage of the specificity of the structure, for which we must pinpoint the most relevant interfacial residues, since they are extremely likely to be crucial for complex formation. Here, we describe step by step our own in silico protocol to characterize these residues, making use of known experimental structures. We detail the computational pipeline from data acquisition and pre-processing to feature extraction. A molecular dynamics simulation protocol to further study membrane dimer proteins and their interfaces is also illustrated.

Background: Selectively targeting dopamine receptors has been a persistent challenge in the last years for the development of new treatments to combat the large variety of diseases evolving these receptors. Although, several drugs have been successfully brought to market, the subtype-specific binding mode on a molecular basis has not been fully elucidated. Methods: Homology modeling and molecular dynamics were applied to construct robust conformational models of all dopamine receptor subtypes (D1-like and D2-like receptors). Fifteen structurally diverse ligands were docked to these models. Contacts at the binding pocket were fully described in order to reveal new structural findings responsible for DR sub-type specificity. Results: We showed that the number of conformations for a receptor:ligand complex was associated to unspecific interactions > 2.5 Å and hydrophobic contacts, while the decoys binding energy was influenced by specific electrostatic interactions. Known residues such as 3.32Asp, the serine microdomain and the aromatic microdomain were found interacting in a variety of modes (HB, SB, π-stacking). Purposed TM2-TM3-TM7 microdomain was found to form a hydrophobic network involving Orthosteric Binding Pocket (OBP) and Secondary Binding Pocket (SBP). T-stacking interactions revealed as especially relevant for some large ligands such as apomorphine, risperidone or aripiprazole. Conclusions: This in silico approach was successful in showing known receptor-ligand interactions as well as in determining unique combinations of interactions, key for the design of more specific ligands.

Background: Selectively targeting dopamine receptors (DRs) has been a persistent challenge in the last years for the development of new treatments to combat the large variety of diseases involving these receptors. Although, several drugs have been successfully brought to market, the subtype-specific binding mode on a molecular basis has not been fully elucidated. Methods: Homology modeling and molecular dynamics were applied to construct robust conformational models of all dopamine receptor subtypes (D1-like and D2-like). Fifteen structurally diverse ligands were docked. Contacts at the binding pocket were fully described in order to reveal new structural findings responsible for selective binding to DR subtypes. Results: Residues of the aromatic microdomain were shown to be responsible for the majority of ligand interactions established to all DRs. Hydrophobic contacts involved a huge network of conserved and non-conserved residues between three transmembrane domains (TMs), TM2-TM3-TM7. Hydrogen bonds were mostly mediated by the serine microdomain. TM1 and TM2 residues were main contributors for the coupling of large ligands. Some amino acid groups form electrostatic interactions of particular importance for D1R-like selective ligands binding. Conclusions: This in silico approach was successful in showing known receptor-ligand interactions as well as in determining unique combinations of interactions, which will support mutagenesis studies to improve the design of subtype-specific ligands.

Membrane proteins are essential vessels for cell communication both with other cells and noncellular structures. They modulate environment responses and mediate a myriad of biological processes. Dimerization and multimerization processes have been shown to further increase the already high specificity of these processes. Due to their central role in various cell and organism functions, these multimers are often associated with health conditions, such as Alzheimer's disease (AD), Parkinson's disease (PD), and diabetes, among others.Understanding the membrane protein dimers' interface takes advantage of the specificity of the structure, for which we must pinpoint the most relevant interfacial residues, since they are extremely likely to be crucial for complex formation. Here, we describe step by step our own in silico protocol to characterize these residues, making use of known experimental structures. We detail the computational pipeline from data acquisition and pre-processing to feature extraction. A molecular dynamics simulation protocol to further study membrane dimer proteins and their interfaces is also illustrated.

Cancer has become one of the main leading causes of morbidity and mortality worldwide. One of the critical drawbacks of current cancer therapeutics has been the lack of the target-selectivity, as these drugs should have an effect exclusively on cancer cells while not perturbing healthy ones. In addition, their mechanism of action should be sufficiently fast to avoid the invasion of neighbouring healthy tissues by cancer cells. The use of conventional chemotherapeutic agents and other traditional therapies, such as surgery and radiotherapy, leads to off-target interactions with serious side effects. In this respect, recently developed target-selective Antibody-Drug Conjugates (ADCs) are more effective than traditional therapies, presumably due to their modular structures that combine many chemical properties simultaneously. In particular, ADCs are made up of three different units: a highly selective Monoclonal antibody (Mab) which is developed against a tumour-associated antigen, the payload (cytotoxic agent), and the linker. The latter should be stable in circulation while allowing the release of the cytotoxic agent in target cells. The modular nature of these drugs provides a platform to manipulate and improve selectivity and the toxicity of these molecules independently from each other. This in turn leads to generation of second- and third-generation ADCs, which have been more effective than the previous ones in terms of either selectivity or toxicity or both. Development of ADCs with improved efficacy requires knowledge at the atomic level regarding the structure and dynamics of the molecule. As such, we reviewed all the most recent computational methods used to attain all-atom description of the structure, energetics and dynamics of these systems. In particular, this includes homology modelling, molecular docking and refinement, atomistic and coarse-grained molecular dynamics simulations, principal component and cross-correlation analysis. The full characterization of the structure-activity relationship devoted to ADCs is critical for antibody-drug conjugate research and development.

Parkinson’s Disease (PD) is a long-term neurodegenerative brain disorder that mainly affects the motor system. The causes are still unknown, and even though currently there is no cure, several therapeutic options are available to manage its symptoms. The development of novel anti-parkinsonian agents and an understanding of their proper and optimal use are, indeed, highly demanding. For the last decades, L-3,4-DihydrOxyPhenylAlanine or levodopa (L-DOPA) has been the gold-standard therapy for the symptomatic treatment of motor dysfunctions associated to PD. However, the development of dyskinesias and motor fluctuations (wearing-off and on-off phenomena) associated with long-term L-DOPA replacement therapy have limited its antiparkinsonian efficacy. The investigation for non-dopaminergic therapies has been largely explored as an attempt to counteract the motor side effects associated with dopamine replacement therapy. Being one of the largest cell membrane protein families, G-Protein-Coupled Receptors (GPCRs) have become a relevant target for drug discovery focused on a wide range of therapeutic areas, including Central Nervous System (CNS) diseases. The modulation of specific GPCRs potentially implicated in PD, excluding dopamine receptors, may provide promising non-dopaminergic therapeutic alternatives for symptomatic treatment of PD. In this review, we focused on the impact of specific GPCR subclasses, including dopamine receptors, adenosine receptors, muscarinic acetylcholine receptors, metabotropic glutamate receptors, and 5-hydroxytryptamine receptors, on the pathophysiology of PD and the importance of structure- and ligand-based in silico approaches for the development of small molecules to target these receptors.

Due to the clinical importance of the Dopamine D2-receptor (D2R) in several brain dysfunctions, the utilization of in silico models for drug development is a growing field of investigation. We provided a transparent and reproducible pipeline for creating a valid D2R model for small molecular docking studies. Furthermore, we suggested a binding pocket for the endogenous ligand of D2R, which was attained upon careful consideration of the available experimental data. Molecular docking studies with Dopamine, Quinpirole and Raclopride allowed also a better understanding of the binding pocket characteristics.

We present SpotOn, a web server to identify and classify interfacial residues as Hot-Spots (HS) and Null-Spots (NS). SpotON implements a robust algorithm with a demonstrated accuracy of 0.95 and sensitivity of 0.98 on an independent test set. The predictor was developed using an ensemble machine learning approach with up-sampling of the minor class. It was trained on 53 complexes using various features, based on both protein 3D structure and sequence. The SpotOn web interface is freely available at: http://milou.science.uu.nl/services/SPOTON/.

Protein-protein interactions (PPIs) have proven necessary for the majority of biological processes, making their understanding vital for the development of new therapies and techniques in life sciences research. Among the residues that constitute a typical protein-protein interface, Hot-Spots (HS) are the most important ones due to their highly stabilizing nature. However, HS experimental detection has proven to be a burden as it is time consuming and expensive, which prompted the need to develop new computational approaches that ensure both speed and precision. Evolution plays a major role in protein structure and PPIs refinement, and therefore the incorporation of such data into a predictive model may lead to better performance. With this in mind and taking into account the data already available from alanine scanning mutagenesis studies and protein structures, we incorporated several structure- (i.e. solvent accessible surface area-related values, sequence- (i.e. position-specific scoring matrix), and evolutionary-based (i.e. InterEVScore and CoeViz) features into a predictive machine-learning classification model. We considered six different pre-processing conditions such as Principal Component Analysis (PCA) and z-scoring (scaling) with normal, up- and down-sampling of minor and major classes. Our results point towards overall better scores when using more evolutionary features, in particular EVFold scores.

Bioinformatics allows to automatically characterize a large number of proteins from numerous different databases, thus, uncovering new possible interactions between biomolecules in a huge set of individuals in a conscious and cost-efficient way. Membrane proteins are indisputably important for the assurance of major processes in the cell, occupying approximately 25% of the whole cell genome. In this work, some of the major features displayed at Protein Data Bank (original species, chains and ligands, oligomer state, multimeric states, stoichiometry, among others) of membrane proteins listed in the Membrane Proteins with Known 3D Structure database were registered together using manual and automated methods - some of these methods include the usage of python specific tools (like Selenium and BioPython). We aimed to construct a membrane-membrane dimers database that will serve as input for data-mining algorithms to unveiling new functional and evolutionary knowledge.

The identification of protein complexes and interactions is crucial for the understanding of cellular organization and machinery. Due to the high difficulty in attaining experimental data about such an important subject, computational tools and methodologies are emerging as reliable alternatives. It is especially true that Machine-Learning (ML) algorithms hold an incredible promise for protein interaction research by identifying biological relevant patterns. In this work, we have developed and applied ML techniques that went beyond the current state-of-the-art, leading to quantitative and reliable molecular-level predictions of Hot-Spots at protein-protein complexes. Our model was trained on a large number of complexes and on a significantly larger number of different structural- and evolutionary sequence-based features. We used 54 algorithms from a simple linear-based function to support-vector machine models with different cost functions and compared their performance in 6 different pre-processing sets. The best model was achieved by the use of the C5.0 tree-based algorithm with a dataset pre-processed by the normalization of features and with up-sampling of the minor class. The method showed an overall accuracy of 0.88, a sensitivity of 0.91 and a specificity of 0.84 for the independent test set. Due to the theme's relevance to the large scientific community working on structural biology, we have assembled a freely available web-server that can be found at: http://milou.science.uu.nl/services/SPOTON/