Abstract: In diabetes, the elevated levels of glucose in the bloodstream can result in the nonenzymatic glycation of proteins such as human serum albumin (HSA). This type of modification has been shown to affect the interactions of some drugs with HSA, including several sulfonylurea drugs that are used to treat type II diabetes. This study used high-performance affinity chromatography (HPAC) to examine the interactions of glipizide (i.e., a second-generation sulfonylurea drug) with normal HSA or HSA that contained various levels of in vitro glycation. Frontal analysis indicated that glipizide was interacting with both normal and glycated HSA through two general groups of sites: a set of relatively strong interactions and a set of weaker interactions with average association equilibrium constants at pH 7.4 and 37 degrees C in the range of 2.4-6.0 x 10(5) and 1.7-3.7 x 10(4) M(-1), respectively. Zonal elution competition studies revealed that glipizide was interacting at both Sudlow sites I and II, which were estimated to have affinities of 3.2-3.9 x 10(5) and 1.1-1.4 x 10(4) M(-1). Allosteric effects were also noted to occur for this drug between the tamoxifen site and the binding of R-warfarin at Sudlow site I. Up to an 18% decrease in the affinity for glipizide was observed at Sudlow site I ongoing from normal HSA to glycated HSA, while up to a 27% increase was noted at Sudlow site II. This information should be useful in indicating how HPAC can be used to investigate other drugs that have complex interactions with proteins. These results should also be valuable in providing a better understanding of how glycation may affect drug-protein interactions and the serum transport of drugs such as glipizide during diabetes.

Abstract: The Psychiatric Genomics Consortium-Schizophrenia Workgroup (PGC-SCZ) has recently published a genomewide association study (GWAS) identifying >100 genetic loci, encompassing a total of 341 protein-coding genes, attaining genomewide significance for susceptibility to schizophrenia. Given the extremely long time (12-15 years) and expense (>$1 billion) associated with the development of novel drug targets, repurposing of drugs with known and validated targets may be the most expeditious path toward deriving clinical utility from these GWAS findings. In the present study, we examined all genes within loci implicated by the PGC-SCZ GWAS against databases of targets of both approved and registered pharmaceutical compounds. We identified 20 potential schizophrenia susceptibility genes that encode proteins that are the targets of approved drugs. Of these, we prioritized genes/targets that are of clear neuropsychiatric interest and that are also sole members of the linkage disequilibrium block surrounding a PGC-SCZ GWAS hit. In addition to DRD2, 5 genes meet these criteria: CACNA1C, CACNB2, CACNA1I, GRIN2A and HCN1. An additional 20 genes coding for proteins that are the targets of drugs in registered clinical trials, but without approved indications, were also identified. Although considerable work is still required to fully explicate the biological implications of the PGC-SCZ GWAS results, pathways related to these known, druggable targets may represent a promising starting point.

Abstract: The molecular mechanism of many drug side-effects is unknown and difficult to predict. Previous methods for explaining side-effects have focused on known drug targets and their pathways. However, low affinity binding to proteins that are not usually considered drug targets may also drive side-effects. In order to assess these alternative targets, we used the 3D structures of 563 essential human proteins systematically to predict binding to 216 drugs. We first benchmarked our affinity predictions with available experimental data. We then combined singular value decomposition and canonical component analysis (SVD-CCA) to predict side-effects based on these novel target profiles. Our method predicts side-effects with good accuracy (average AUC: 0.82 for side effects present in <50% of drug labels). We also noted that side-effect frequency is the most important feature for prediction and can confound efforts at elucidating mechanism; our method allows us to remove the contribution of frequency and isolate novel biological signals. In particular, our analysis produces 2768 triplet associations between 50 essential proteins, 99 drugs, and 77 side-effects. Although experimental validation is difficult because many of our essential proteins do not have validated assays, we nevertheless attempted to validate a subset of these associations using experimental assay data. Our focus on essential proteins allows us to find potential associations that would likely be missed if we used recognized drug targets. Our associations provide novel insights about the molecular mechanisms of drug side-effects and highlight the need for expanded experimental efforts to investigate drug binding to proteins more broadly.

Abstract: Large-scale databases are important sources of training sets for various QSAR modeling approaches. Generally, these databases contain information extracted from different sources. This variety of sources can produce inconsistency in the data, defined as sometimes widely diverging activity results for the same compound against the same target. Because such inconsistency can reduce the accuracy of predictive models built from these data, we are addressing the question of how best to use data from publicly and commercially accessible databases to create accurate and predictive QSAR models. We investigate the suitability of commercially and publicly available databases to QSAR modeling of antiviral activity (HIV-1 reverse transcriptase (RT) inhibition). We present several methods for the creation of modeling (i.e., training and test) sets from two, either commercially or freely available, databases: Thomson Reuters Integrity and ChEMBL. We found that the typical predictivities of QSAR models obtained using these different modeling set compilation methods differ significantly from each other. The best results were obtained using training sets compiled for compounds tested using only one method and material (i.e., a specific type of biological assay). Compound sets aggregated by target only typically yielded poorly predictive models. We discuss the possibility of "mix-and-matching" assay data across aggregating databases such as ChEMBL and Integrity and their current severe limitations for this purpose. One of them is the general lack of complete and semantic/computer-parsable descriptions of assay methodology carried by these databases that would allow one to determine mix-and-matchability of result sets at the assay level.

Abstract: Pharmaceutical research has successfully incorporated a wealth of molecular modeling methods, within a variety of drug discovery programs, to study complex biological and chemical systems. The integration of computational and experimental strategies has been of great value in the identification and development of novel promising compounds. Broadly used in modern drug design, molecular docking methods explore the ligand conformations adopted within the binding sites of macromolecular targets. This approach also estimates the ligand-receptor binding free energy by evaluating critical phenomena involved in the intermolecular recognition process. Today, as a variety of docking algorithms are available, an understanding of the advantages and limitations of each method is of fundamental importance in the development of effective strategies and the generation of relevant results. The purpose of this review is to examine current molecular docking strategies used in drug discovery and medicinal chemistry, exploring the advances in the field and the role played by the integration of structure- and ligand-based methods.

Abstract: Drug-drug interaction (DDI) is an important topic for public health, and thus attracts attention from both academia and industry. Here we hypothesize that clinical side effects (SEs) provide a human phenotypic profile and can be translated into the development of computational models for predicting adverse DDIs. We propose an integrative label propagation framework to predict DDIs by integrating SEs extracted from package inserts of prescription drugs, SEs extracted from FDA Adverse Event Reporting System, and chemical structures from PubChem. Experimental results based on hold-out validation demonstrated the effectiveness of the proposed algorithm. In addition, the new algorithm also ranked drug information sources based on their contributions to the prediction, thus not only confirming that SEs are important features for DDI prediction but also paving the way for building more reliable DDI prediction models by prioritizing multiple data sources. By applying the proposed algorithm to 1,626 small-molecule drugs which have one or more SE profiles, we obtained 145,068 predicted DDIs. The predicted DDIs will help clinicians to avoid hazardous drug interactions in their prescriptions and will aid pharmaceutical companies to design large-scale clinical trial by assessing potentially hazardous drug combinations. All data sets and predicted DDIs are available at http://astro.temple.edu/~tua87106/ddi.html.

Abstract: Genome-wide identification of the mechanism of action (MoA) of small-molecule compounds characterizing their targets, effectors, and activity modulators represents a highly relevant yet elusive goal, with critical implications for assessment of compound efficacy and toxicity. Current approaches are labor intensive and mostly limited to elucidating high-affinity binding target proteins. We introduce a regulatory network-based approach that elucidates genome-wide MoA proteins based on the assessment of the global dysregulation of their molecular interactions following compound perturbation. Analysis of cellular perturbation profiles identified established MoA proteins for 70% of the tested compounds and elucidated novel proteins that were experimentally validated. Finally, unknown-MoA compound analysis revealed altretamine, an anticancer drug, as an inhibitor of glutathione peroxidase 4 lipid repair activity, which was experimentally confirmed, thus revealing unexpected similarity to the activity of sulfasalazine. This suggests that regulatory network analysis can provide valuable mechanistic insight into the elucidation of small-molecule MoA and compound similarity.

Abstract: The metabotropic glutamate subtype 1 (mGluR1), a member of the metabotropic glutamate receptors, is a therapeutic target for neurological disorders. However, due to the lower subtype selectivity of mGluR1 orthosteric compounds, a new targeted strategy, known as allosteric modulators research, is needed for the treatment of mGluR1-related diseases. Recently, the structure of the seven-transmembrane domain (7TMD) of mGluR1 has been solved, which reveals the binding site of allosteric modulators and provides an opportunity for future subtype-selectivity drug design. In this study, a series of computer-aided drug design methods were utilized to discover potential mGluR1 negative allosteric modulators (NAMs). Pharmacophore models were constructed based on three different structure types of mGluR1 NAMs. After validation using the built-in parameters and test set, the optimal pharmacophore model of each structure type was selected and utilized as a query to screen the Traditional Chinese Medicine Database (TCMD). Then, three different hit lists of compounds were obtained. Molecular docking was used based on the latest crystal structure of mGluR1-7TMD to further filter these hits. As a compound with high QFIT and LibDock Score was preferred, a total of 30 compounds were retained. MD simulation was utilized to confirm the stability of potential compounds binding. From the computational results, thesinine-4'-O-beta-d-glucoside, nigrolineaxanthone-P and nodakenin might exhibit negative allosteric moderating effects on mGluR1. This paper indicates the applicability of molecular simulation technologies for discovering potential natural mGluR1 NAMs from Chinese herbs.

Abstract: Transcription factors (TFs) and microRNAs (miRNAs) form a gene regulatory network (GRN) at the transcriptional and post-transcriptional level in living cells. However, this network has not been well characterized, especially in regards to the mutual regulations between TFs and miRNAs in cancers. In this study, we collected those regulations inferred by ChIP-Seq or CLIP-Seq to construct the GRN formed by TFs, miRNAs, and target genes. To increase the reliability of the proposed network and examine the regulation activity of TFs and miRNAs, we further incorporated the mRNA and miRNA expression profiles in seven cancer types using The Cancer Genome Atlas data. We observed that regulation rewiring was prevalent during tumorigenesis and found that the rewired regulatory feedback loops formed by TFs and miRNAs were highly associated with cancer. Interestingly, we identified one regulatory feedback loop between STAT1 and miR-155-5p that is consistently activated in all seven cancer types with its function to regulate tumor-related biological processes. Our results provide insights on the losing equilibrium of the regulatory feedback loop between STAT1 and miR-155-5p influencing tumorigenesis.

Abstract: BACKGROUND: Active learning is a powerful tool for guiding an experimentation process. Instead of doing all possible experiments in a given domain, active learning can be used to pick the experiments that will add the most knowledge to the current model. Especially, for drug discovery and development, active learning has been shown to reduce the number of experiments needed to obtain high-confidence predictions. However, in practice, it is crucial to have a method to evaluate the quality of the current predictions and decide when to stop the experimentation process. Only by applying reliable stopping criteria to active learning can time and costs in the experimental process actually be saved. RESULTS: We compute active learning traces on simulated drug-target matrices in order to determine a regression model for the accuracy of the active learner. By analyzing the performance of the regression model on simulated data, we design stopping criteria for previously unseen experimental matrices. We demonstrate on four previously characterized drug effect data sets that applying the stopping criteria can result in upto 40 % savings of the total experiments for highly accurate predictions. CONCLUSIONS: We show that active learning accuracy can be predicted using simulated data and results in substantial savings in the number of experiments required to make accurate drug-target predictions.