One of the important questions in aging research is how differences in transcriptomics are associated with the longevity of various species. Unfortunately, at the level of individual genes, the links between expression in different organs and maximum lifespan (MLS) are yet to be fully understood. Analyses are complicated further by the fact that MLS is highly associated with other confounding factors (metabolic rate, gestation period, body mass, etc.) and that linear models may be limiting. Using gene expression from 41 mammalian species, across five organs, we constructed gene-centric regression models associating gene expression with MLS and other species traits. Additionally, we used SHapley Additive exPlanations and Bayesian networks to investigate the non-linear nature of the interrelations between the genes predicted to be determinants of species MLS. Our results revealed that expression patterns correlate with MLS, some across organs, and others in an organ-specific manner. The combination of methods employed revealed gene signatures formed by only a few genes that are highly predictive towards MLS, which could be used to identify novel longevity regulator candidates in mammals.

Gaining insight into the mechanisms of signal transduction networks (STNs) by using critical features from patient-specific mathematical models can improve patient stratification and help to identify potential drug targets. To achieve this, these models should focus on the critical STNs for each cancer, include prognostic genes and proteins, and correctly predict patient-specific differences in STN activity. Focussing on colorectal cancer and the WNT STN, we used mechanism-based machine learning models to identify genes and proteins with significant associations to event-free patient survival and predictive power for explaining patient-specific differences of STN activity. First, we identified the WNT pathway as the most significant pathway associated with event-free survival. Second, we built linear-regression models that incorporated both genes and proteins from established mechanistic models in the literature and novel genes with significant associations to event-free patient survival. Data from The Cancer Genome Atlas and Clinical Proteomic Tumour Analysis Consortium were used, and patient-specific STN activity scores were computed using PROGENy. Three linear regression models were built, based on; (1) the gene-set of a state-of-the-art mechanistic model in the literature, (2) novel genes identified, and (3) novel proteins identified. The novel genes and proteins were genes and proteins of the extant WNT pathway whose expression was significantly associated with event-free survival. The results show that the predictive power of a model that incorporated novel event-free associated genes is better compared to a model focussing on the genes of a current state-of-the-art mechanistic model. Several significant genes that should be integrated into future mechanistic models of the WNT pathway are DVL3, FZD5, RAC1, ROCK2, GSK3B, CTB2, CBT1, and PRKCA. Thus, the study demonstrates that using mechanistic information in combination with machine learning can identify novel features (genes and proteins) that are important for explaining the STN heterogeneity between patients and their association to clinical outcomes.

Cyclin-dependent kinase 5 (Cdk5) is an atypical proline-directed serine/threonine protein kinase well-characterized for its role in the central nervous system rather than in the cell cycle. Indeed, its dysregulation has been strongly implicated in the progression of synaptic dysfunction and neurodegenerative diseases, such as Alzheimer's disease (AD) and Parkinson's disease (PD), and also in the development and progression of a variety of cancers. For this reason, Cdk5 is considered as a promising target for drug design, and the discovery of novel small-molecule Cdk5 inhibitors is of great interest in the medicinal chemistry field. In this context, we employed a machine learning-based virtual screening protocol with subsequent molecular docking, molecular dynamics simulations and binding free energy evaluations. Our virtual screening studies resulted in the identification of two novel Cdk5 inhibitors, highlighting an experimental hit rate of 50% and thus validating the reliability of the in silico workflow. Both identified ligands, compounds CPD1 and CPD4, showed a promising enzyme inhibitory activity and CPD1 also demonstrated a remarkable antiproliferative activity in ovarian and colon cancer cells. These ligands represent a valuable starting point for structure-based hit-optimization studies aimed at identifying new potent Cdk5 inhibitors.

The expression and regulation of matrisome genes-the ensemble of extracellular matrix, ECM, ECM-associated proteins and regulators as well as cytokines, chemokines and growth factors-is of paramount importance for many biological processes and signals within the tumor microenvironment. The availability of large and diverse multi-omics data enables mapping and understanding of the regulatory circuitry governing the tumor matrisome to an unprecedented level, though such a volume of information requires robust approaches to data analysis and integration. In this study, we show that combining Pan-Cancer expression data from The Cancer Genome Atlas (TCGA) with genomics, epigenomics and microenvironmental features from TCGA and other sources enables the identification of "landmark" matrisome genes and machine learning-based reconstruction of their regulatory networks in 74 clinical and molecular subtypes of human cancers and approx. 6700 patients. These results, enriched for prognostic genes and cross-validated markers at the protein level, unravel the role of genetic and epigenetic programs in governing the tumor matrisome and allow the prioritization of tumor-specific matrisome genes (and their regulators) for the development of novel therapeutic approaches.

Identification of the protein targets of hit molecules is essential in the drug discovery process. Target prediction with machine learning algorithms can help accelerate this search, limiting the number of required experiments. However, Drug-Target Interactions databases used for training present high statistical bias, leading to a high number of false positives, thus increasing time and cost of experimental validation campaigns. To minimize the number of false positives among predicted targets, we propose a new scheme for choosing negative examples, so that each protein and each drug appears an equal number of times in positive and negative examples. We artificially reproduce the process of target identification for three specific drugs, and more globally for 200 approved drugs. For the detailed three drug examples, and for the larger set of 200 drugs, training with the proposed scheme for the choice of negative examples improved target prediction results: the average number of false positives among the top ranked predicted targets decreased, and overall, the rank of the true targets was improved.Our method corrects databases' statistical bias and reduces the number of false positive predictions, and therefore the number of useless experiments potentially undertaken.

. De novo drug design is a computational approach that generates novel molecular structures from atomic building blocks with no a priori relationships. Conventional methods include structure-based and ligand-based design, which depend on the properties of the active site of a biological target or its known active binders, respectively. Artificial intelligence, including machine learning, is an emerging field that has positively impacted the drug discovery process. Deep reinforcement learning is a subdivision of machine learning that combines artificial neural networks with reinforcement-learning architectures. This method has successfully been employed to develop novel de novo drug design approaches using a variety of artificial networks including recurrent neural networks, convolutional neural networks, generative adversarial networks, and autoencoders. This review article summarizes advances in de novo drug design, from conventional growth algorithms to advanced machine-learning methodologies and highlights hot topics for further development.

The requirement of innovative big data analytics has become a critical success factor for research in biological psychiatry. Integrative analyses across distributed data resources are considered essential for untangling the biological complexity of mental illnesses. However, little is known about algorithm properties for such integrative machine learning. Here, we performed a comparative analysis of eight machine learning algorithms for identification of reproducible biological fingerprints across data sources, using five transcriptome-wide expression datasets of schizophrenia patients and controls as a use case. We found that multi-task learning (MTL) with network structure (MTL_NET) showed superior accuracy compared to other MTL formulations as well as single task learning, and tied performance with support vector machines (SVM). Compared to SVM, MTL_NET showed significant benefits regarding the variability of accuracy estimates, as well as its robustness to cross-dataset and sampling variability. These results support the utility of this algorithm as a flexible tool for integrative machine learning in psychiatry.

Understanding protein-protein interactions is a key challenge in biochemistry. In this work, we describe a more accurate methodology to predict Hot-Spots (HS) in protein-protein interfaces from their native complex structure compared to previous published Machine Learning (ML) techniques. Our model is trained on a large number of complexes and on a significantly larger number of different structural- and evolutionary sequence-based features. In particular, we added interface size, type of interaction between residues at the interface of the complex, number of different types of residues at the interface and the Position-Specific Scoring Matrix (PSSM), for a total of 79 features. We used twenty-seven algorithms from a simple linear-based function to support-vector machine models with different cost functions. The best model was achieved by the use of the conditional inference random forest (c-forest) algorithm with a dataset pre-processed by the normalization of features and with up-sampling of the minor class. The method has an overall accuracy of 0.80, an F1-score of 0.73, a sensitivity of 0.76 and a specificity of 0.82 for the independent test set.

Cardiac amyloidosis is an uncommon restrictive cardiomyopathy featuring an unregulated amyloid protein deposition that impairs organic function. Early cardiac amyloidosis diagnosis is generally delayed by indistinguishable clinical findings of more frequent hypertrophic diseases. Furthermore, amyloidosis is divided into various groups, according to a generally accepted taxonomy, based on the proteins that make up the amyloid deposits; a careful differentiation between the various forms of amyloidosis is necessary to undertake an adequate therapeutic treatment. Thus, cardiac amyloidosis is thought to be underdiagnosed, which delays necessary therapeutic procedures, diminishing quality of life and impairing clinical prognosis. The diagnostic work-up for cardiac amyloidosis begins with the identification of clinical features, electrocardiographic and imaging findings suggestive or compatible with cardiac amyloidosis, and often requires the histological demonstration of amyloid deposition. One approach to overcome the difficulty of an early diagnosis is the use of automated diagnostic algorithms. Machine learning enables the automatic extraction of salient information from "raw data" without the need for pre-processing methods based on the a priori knowledge of the human operator. This review attempts to assess the various diagnostic approaches and artificial intelligence computational techniques in the detection of cardiac amyloidosis.

Monoclonal antibodies are biopharmaceuticals with a very long half-life due to the binding of their Fc portion to the neonatal receptor (FcRn), a pharmacokinetic property that can be further improved through engineering of the Fc portion, as demonstrated by the approval of several new drugs. Many Fc variants with increased binding to FcRn have been found using different methods, such as structure-guided design, random mutagenesis, or a combination of both, and are described in the literature as well as in patents. Our hypothesis is that this material could be subjected to a machine learning approach in order to generate new variants with similar properties. We therefore compiled 1323 Fc variants affecting the affinity for FcRn, which were disclosed in twenty patents. These data were used to train several algorithms, with two different models, in order to predict the affinity for FcRn of new randomly generated Fc variants. To determine which algorithm was the most robust, we first assessed the correlation between measured and predicted affinity in a 10-fold cross-validation test. We then generated variants by in silico random mutagenesis and compared the prediction made by the different algorithms. As a final validation, we produced variants, not described in any patents, and compared the predicted affinity with the experimental binding affinities measured by surface plasmon resonance (SPR). The best mean absolute error (MAE) between predicted and experimental values was obtained with a support vector regressor (SVR) using six features and trained on 1251 examples. With this setting, the error on the log(KD) was less than 0.17. The obtained results show that such an approach could be used to find new variants with better half-life properties that are different from those already extensively used in therapeutic antibody development.

Experimental studies of Caenorhabditis elegans and Drosophila melanogaster have contributed substantially to our understanding of molecular and cellular processes in metazoans at large. Since the publication of their genomes, functional genomic investigations have identified genes that are essential or non-essential for survival in each species. Recently, a range of features linked to gene essentiality have been inferred using a machine learning (ML)-based approach, allowing essentiality predictions within a species. Nevertheless, predictions between species are still elusive. Here, we undertake a comprehensive study using ML to discover and validate features of essential genes common to both C. elegans and D. melanogaster. We demonstrate that the cross-species prediction of gene essentiality is possible using a subset of features linked to nucleotide/protein sequences, protein orthology and subcellular localisation, single-cell RNA-seq, and histone methylation markers. Complementary analyses showed that essential genes are enriched for transcription and translation functions and are preferentially located away from heterochromatin regions of C. elegans and D. melanogaster chromosomes. The present work should enable the cross-prediction of essential genes between model and non-model metazoans.

The standard approach to genetic mapping was supplemented by machine learning (ML) to establish the location of the rye gene associated with epicuticular wax formation (glaucous phenotype). Over 180 plants of the biparental F2 population were genotyped with the DArTseq (sequencing-based diversity array technology). A maximum likelihood (MLH) algorithm (JoinMap 5.0) and three ML algorithms: logistic regression (LR), random forest and extreme gradient boosted trees (XGBoost), were used to select markers closely linked to the gene encoding wax layer. The allele conditioning the nonglaucous appearance of plants, derived from the cultivar Karlikovaja Zelenostebelnaja, was mapped at the chromosome 2R, which is the first report on this localization. The DNA sequence of DArT-Silico 3585843, closely linked to wax segregation detected by using ML methods, was indicated as one of the candidates controlling the studied trait. The putative gene encodes the ABCG11 transporter.

The large amount of data that has been collected so far for G protein-coupled receptors requires machine learning (ML) approaches to fully exploit its potential. Our previous ML model based on gradient boosting used for prediction of drug affinity and selectivity for a receptor subtype was compared with explicit information on ligand-receptor interactions from induced-fit docking. Both methods have proved their usefulness in drug response predictions. Yet, their successful combination still requires allosteric/orthosteric assignment of ligands from datasets. Our ligand datasets included activities of two members of the secretin receptor family: GCGR and GLP-1R. Simultaneous activation of two or three receptors of this family by dual or triple agonists is not a typical kind of information included in compound databases. A precise allosteric/orthosteric ligand assignment requires a continuous update based on new structural and biological data. This data incompleteness remains the main obstacle for current ML methods applied to class B GPCR drug discovery. Even so, for these two class B receptors, our ligand-based ML model demonstrated high accuracy (5-fold cross-validation Q2 > 0.63 and Q2 > 0.67 for GLP-1R and GCGR, respectively). In addition, we performed a ligand annotation using recent cryogenic-electron microscopy (cryo-EM) and X-ray crystallographic data on small-molecule complexes of GCGR and GLP-1R. As a result, we assigned GLP-1R and GCGR actives deposited in ChEMBL to four small-molecule binding sites occupied by positive and negative allosteric modulators and a full agonist. Annotated compounds were added to our recently released repository of GPCR data.

Psoriasis (Pso) is a chronic inflammatory skin disease, and up to 30% of Pso patients develop psoriatic arthritis (PsA), which can lead to irreversible joint damage. Early detection of PsA in Pso patients is crucial for timely treatment but difficult for dermatologists to implement. We, therefore, aimed to find disease-specific immune profiles, discriminating Pso from PsA patients, possibly facilitating the correct identification of Pso patients in need of referral to a rheumatology clinic. The phenotypes of peripheral blood immune cells of consecutive Pso and PsA patients were analyzed, and disease-specific immune profiles were identified via a machine learning approach. This approach resulted in a random forest classification model capable of distinguishing PsA from Pso (mean AUC = 0.95). Key PsA-classifying cell subsets selected included increased proportions of differentiated CD4+CD196+CD183-CD194+ and CD4+CD196-CD183-CD194+ T-cells and reduced proportions of CD196+ and CD197+ monocytes, memory CD4+ and CD8+ T-cell subsets and CD4+ regulatory T-cells. Within PsA, joint scores showed an association with memory CD8+CD45RA-CD197- effector T-cells and CD197+ monocytes. To conclude, through the integration of in-depth flow cytometry and machine learning, we identified an immune cell profile discriminating PsA from Pso. This immune profile may aid in timely diagnosing PsA in Pso.

This paper describes a machine learning (ML) decision support system to provide a list of chemotherapeutics that individual multiple myeloma (MM) patients are sensitive/resistant to, based on their proteomic profile. The methodology used in this study involved understanding the parameter space and selecting the dominant features (proteomics data), identifying patterns of proteomic profiles and their association to the recommended treatments, and defining the decision support system of personalized treatment as a classification problem. During the data analysis, we compared several ML algorithms, such as linear regression, Random Forest, and support vector machines, to classify patients as sensitive/resistant to therapeutics. A further analysis examined data-balancing techniques that emerged due to the small cohort size. The results suggest that utilizing proteomics data is a promising approach for identifying effective treatment options for patients with MM (reaching on average an accuracy of 81%). Although this pilot study was limited by the small patient cohort (39 patients), which restricted the training and validation of the explored ML solutions to identify complex associations between proteins, it holds great promise for developing personalized anti-MM treatments using ML approaches.

Olfactory receptors (ORs) constitute the largest superfamily of G protein-coupled receptors (GPCRs). ORs are involved in sensing odorants as well as in other ectopic roles in non-nasal tissues. Matching of an enormous number of the olfactory stimulation repertoire to its counterpart OR through machine learning (ML) will enable understanding of olfactory system, receptor characterization, and exploitation of their therapeutic potential. In the current study, we have selected two broadly tuned ectopic human OR proteins, OR1A1 and OR2W1, for expanding their known chemical space by using molecular descriptors. We present a scheme for selecting the optimal features required to train an ML-based model, based on which we selected the random forest (RF) as the best performer. High activity agonist prediction involved screening five databases comprising ~23 M compounds, using the trained RF classifier. To evaluate the effectiveness of the machine learning based virtual screening and check receptor binding site compatibility, we used docking of the top target ligands to carefully develop receptor model structures. Finally, experimental validation of selected compounds with significant docking scores through in vitro assays revealed two high activity novel agonists for OR1A1 and one for OR2W1.

Glucose-6-Phosphate Dehydrogenase (G6PD) is a ubiquitous cytoplasmic enzyme converting glucose-6-phosphate into 6-phosphogluconate in the pentose phosphate pathway (PPP). The G6PD deficiency renders the inability to regenerate glutathione due to lack of Nicotine Adenosine Dinucleotide Phosphate (NADPH) and produces stress conditions that can cause oxidative injury to photoreceptors, retinal cells, and blood barrier function. In this study, we constructed pharmacophore-based models based on the complex of G6PD with compound AG1 (G6PD activator) followed by virtual screening. Fifty-three hit molecules were mapped with core pharmacophore features. We performed molecular descriptor calculation, clustering, and principal component analysis (PCA) to pharmacophore hit molecules and further applied statistical machine learning methods. Optimal performance of pharmacophore modeling and machine learning approaches classified the 53 hits as drug-like (18) and nondrug-like (35) compounds. The drug-like compounds further evaluated our established cheminformatics pipeline (molecular docking and in silico ADMET (absorption, distribution, metabolism, excretion and toxicity) analysis). Finally, five lead molecules with different scaffolds were selected by binding energies and in silico ADMET properties. This study proposes that the combination of machine learning methods with traditional structure-based virtual screening can effectively strengthen the ability to find potential G6PD activators used for G6PD deficiency diseases. Moreover, these compounds can be considered as safe agents for further validation studies at the cell level, animal model, and even clinic setting.

The theoretical prediction of drug-decorated nanoparticles (DDNPs) has become a very important task in medical applications. For the current paper, Perturbation Theory Machine Learning (PTML) models were built to predict the probability of different pairs of drugs and nanoparticles creating DDNP complexes with anti-glioblastoma activity. PTML models use the perturbations of molecular descriptors of drugs and nanoparticles as inputs in experimental conditions. The raw dataset was obtained by mixing the nanoparticle experimental data with drug assays from the ChEMBL database. Ten types of machine learning methods have been tested. Only 41 features have been selected for 855,129 drug-nanoparticle complexes. The best model was obtained with the Bagging classifier, an ensemble meta-estimator based on 20 decision trees, with an area under the receiver operating characteristic curve (AUROC) of 0.96, and an accuracy of 87% (test subset). This model could be useful for the virtual screening of nanoparticle-drug complexes in glioblastoma. All the calculations can be reproduced with the datasets and python scripts, which are freely available as a GitHub repository from authors.

Although extensive advancements have been made in treatment against hepatocellular carcinoma (HCC), the prognosis of HCC patients remains unsatisfied. It is now clearly established that extensive epigenetic changes act as a driver in human tumors. This study exploits HCC epigenetic deregulation to define a novel prognostic model for monitoring the progression of HCC. We analyzed the genome-wide DNA methylation profile of 374 primary tumor specimens using the Illumina 450 K array data from The Cancer Genome Atlas. We initially used a novel combination of Machine Learning algorithms (Recursive Features Selection, Boruta) to capture early tumor progression features. The subsets of probes obtained were used to train and validate Random Forest models to predict a Progression Free Survival greater or less than 6 months. The model based on 34 epigenetic probes showed the best performance, scoring 0.80 accuracy and 0.51 Matthews Correlation Coefficient on testset. Then, we generated and validated a progression signature based on 4 methylation probes capable of stratifying HCC patients at high and low risk of progression. Survival analysis showed that high risk patients are characterized by a poorer progression free survival compared to low risk patients. Moreover, decision curve analysis confirmed the strength of this predictive tool over conventional clinical parameters. Functional enrichment analysis highlighted that high risk patients differentiated themselves by the upregulation of proliferative pathways. Ultimately, we propose the oncogenic MCM2 gene as a methylation-driven gene of which the representative epigenetic markers could serve both as predictive and prognostic markers. Briefly, our work provides several potential HCC progression epigenetic biomarkers as well as a new signature that may enhance patients surveillance and advances in personalized treatment.

Cardiovascular diseases (CDs) are a major concern in the human race and one of the leading causes of death worldwide. β-Adrenergic receptors (β1-AR and β2-AR) play a crucial role in the overall regulation of cardiac function. In the present study, structure-based virtual screening, machine learning (ML), and a ligand-based similarity search were conducted for the PubChem database against both β1- and β2-AR. Initially, all docked molecules were screened using the threshold binding energy value. Molecules with a better binding affinity were further used for segregation as active and inactive through ML. The pharmacokinetic assessment was carried out on molecules retained in the above step. Further, similarity searching of the ChEMBL and DrugBank databases was performed. From detailed analysis of the above data, four compounds for each of β1- and β2-AR were found to be promising in nature. A number of critical ligand-binding amino acids formed potential hydrogen bonds and hydrophobic interactions. Finally, a molecular dynamics (MD) simulation study of each molecule bound with the respective target was performed. A number of parameters obtained from the MD simulation trajectories were calculated and substantiated the stability between the protein-ligand complex. Hence, it can be postulated that the final molecules might be crucial for CDs subjected to experimental validation.