LINCS Canvas Browser: interactive web app to query, browse and interrogate LINCS L1000 gene expression signatures

De Novo Generation of Chemical Structures of Inhibitor and Activator Candidates for Therapeutic Target Proteins by a Transformer-Based Variational Autoencoder and Bayesian Optimization

Deep generative models for molecular generation have been gaining much attention as structure generators to accelerate drug discovery. However, most previously developed methods are chemistry-centric approaches, and comprehensive biological responses in the cell have not been taken into account. In this study, we propose a novel computational method, TRIOMPHE-BOA (transcriptome-based inference and generation of molecules with desired phenotypes using the Bayesian optimization algorithm), to generate new chemical structures of inhibitor or activator candidates for therapeutic target proteins by integrating chemically and genetically perturbed transcriptome profiles. In the algorithm, the substructures of multiple molecules that were selected based on the transcriptome analysis are fused in the design of new chemical structures by exploring the latent space of a Transformer-based variational autoencoder using Bayesian optimization. Our results demonstrate the usefulness of the proposed method in terms of having high reproducibility of existing ligands for 10 therapeutic target proteins when compared with previous methods. Moreover, this method can be applied to proteins without detailed 3D structures or known ligands and is expected to become a powerful tool for more efficient hit identification.

Drug repurposing based on differentially expressed genes suggests drug combinations with possible synergistic effects in treatment of lung adenocarcinoma

Lung adenocarcinoma is one of the leading causes of cancer-related mortality globally. Various treatment approaches and drugs had little influence on overall survival; thus, new drugs and treatment strategies are needed. Drug repositioning (repurposing) seems a favorable approach considering that developing new drugs needs much more time and costs. We performed a meta-analysis on 6 microarray datasets to obtain the main genes with significantly altered expression in lung adenocarcinoma. Following that, we found major gene clusters and hub genes. We assessed their enrichment in biological pathways to get insight into the underlying biological process involved in lung adenocarcinoma pathogenesis. The L1000 database was explored for drug perturbations that might reverse the expression of differentially expressed genes in lung adenocarcinoma. We evaluated the potential drug combinations that interact the most with hub genes and hence have the most potential to reverse the disease process. A total of 2148 differentially expressed genes were identified. Six main gene clusters and 27 significant hub genes mainly involved in cell cycle regulation have been identified. By assessing the interaction between 3 drugs and hub genes and information gained from previous clinical investigations, we suggested the three possible repurposed drug combinations, Vorinostat - Dorsomorphin, PP-110 - Dorsomorphin, and Puromycin - Vorinostat with a high chance of success in clinical trials.

Exogenous Chemicals Impact Virus Receptor Gene Transcription: Insights from Deep Learning

Despite the fact that coronavirus disease 2019 (COVID-19), caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), has been disrupting human life and health worldwide since the outbreak in late 2019, the impact of exogenous substance exposure on the viral infection remains unclear. It is well-known that, during viral infection, organism receptors play a significant role in mediating the entry of viruses to enter host cells. A major receptor of SARS-CoV-2 is the angiotensin-converting enzyme 2 (ACE2). This study proposes a deep learning model based on the graph convolutional network (GCN) that enables, for the first time, the prediction of exogenous substances that affect the transcriptional expression of the ACE2 gene. It outperforms other machine learning models, achieving an area under receiver operating characteristic curve (AUROC) of 0.712 and 0.703 on the validation and internal test set, respectively. In addition, quantitative polymerase chain reaction (qPCR) experiments provided additional supporting evidence for indoor air pollutants identified by the GCN model. More broadly, the proposed methodology can be applied to predict the effect of environmental chemicals on the gene transcription of other virus receptors as well. In contrast to typical deep learning models that are of black box nature, we further highlight the interpretability of the proposed GCN model and how it facilitates deeper understanding of gene change at the structural level.

Applying a Gene Reversal Rate Computational Methodology to Identify Drugs for a Rare Cancer: Inflammatory Breast Cancer

The aim of this study was to utilize a computational methodology based on Gene Reversal Rate (GRR) scoring to repurpose existing drugs for a rare and understudied cancer: inflammatory breast cancer (IBC). This method uses IBC-related gene expression signatures (GES) and drug-induced gene expression profiles from the LINCS database to calculate a GRR score for each candidate drug, and is based on the idea that a compound that can counteract gene expression changes of a disease may have potential therapeutic applications for that disease. Genes related to IBC with associated differential expression data (265 up-regulated and 122 down-regulated) were collated from PubMed-indexed publications. Drug-induced gene expression profiles were downloaded from the LINCS database and candidate drugs to treat IBC were predicted using their GRR scores. Thirty-two (32) drug perturbations that could potentially reverse the pre-compiled list of 297 IBC genes were obtained using the LINCS Canvas Browser (LCB) analysis. Binary combinations of the 32 perturbations were assessed computationally to identify combined perturbations with the highest GRR scores, and resulted in 131 combinations with GRR greater than 80%, that reverse up to 264 of the 297 genes in the IBC-GES. The top 35 combinations involve 20 unique individual drug perturbations, and 19 potential drug candidates. A comprehensive literature search confirmed 17 of the 19 known drugs as having either anti-cancer or anti-inflammatory activities. AZD-7545, BMS-754807, and nimesulide target known IBC relevant genes: PDK, Met, and COX, respectively. AG-14361, butalbital, and clobenpropit are known to be functionally relevant in DNA damage, cell cycle, and apoptosis, respectively. These findings support the use of the GRR approach to identify drug candidates and potential combination therapies that could be used to treat rare diseases such as IBC.

Development of a MEK inhibitor, NFX-179, as a chemoprevention agent for squamous cell carcinoma

Cutaneous squamous cell carcinoma (cSCC) is the second most common skin cancer. Although cSCC contributes to substantial morbidity and mortality in high-risk individuals, deployment of otherwise effective chemoprevention of cSCC is limited by toxicities. Our systematic computational drug repurposing screen predicted that selumetinib, a MAPK (mitogen-activated protein kinase) kinase inhibitor (MEKi), would reverse transcriptional signatures associated with cSCC development, consistent with our genomic analysis implicating MEK as a chemoprevention target. Although systemic MEKi suppresses the formation of cSCC in mice, systemic MEKi can cause severe adverse effects. Here, we report the development of a metabolically labile MEKi, NFX-179, designed to potently and selectively suppress the MAPK pathway in the skin before rapid metabolism in the systemic circulation. NFX-179 was identified on the basis of its biochemical and cellular potency, selectivity, and rapid metabolism upon systemic absorption. In our ultraviolet-induced cSCC mouse model, topical application of NFX-179 gel reduced the formation of new cSCCs by an average of 60% at doses of 0.1% and greater at 28 days. We further confirmed the localized nature of these effects in an additional split-mouse randomized controlled study where suppression of cSCC was observed only in drug-treated areas. No toxicities were observed. NFX-179 inhibits the growth of human SCC cell lines in a dose-dependent manner, and topical NFX-179 application penetrates human skin and inhibits MAPK signaling in human cSCC explants. Together, our data provide a compelling rationale for using topical MEK inhibition through the application of NFX-179 gel as an effective strategy for cSCC chemoprevention.

Predicting synergistic anticancer drug combination based on low-rank global attention mechanism and bilinear predictor

MOTIVATION

Drug combination therapy has exhibited remarkable therapeutic efficacy and has gradually become a promising clinical treatment strategy of complex diseases such as cancers. As the related databases keep expanding, computational methods based on deep learning model have become powerful tools to predict synergistic drug combinations. However, predicting effective synergistic drug combinations is still a challenge due to the high complexity of drug combinations, the lack of biological interpretability, and the large discrepancy in the response of drug combinations in vivo and in vitro biological systems.

RESULTS

Here, we propose DGSSynADR, a new deep learning method based on global structured features of drugs and targets for predicting synergistic anticancer drug combinations. DGSSynADR constructs a heterogeneous graph by integrating the drug-drug, drug-target, protein-protein interactions and multi-omics data, utilizes a low-rank global attention (LRGA) model to perform global weighted aggregation of graph nodes and learn the global structured features of drugs and targets, and then feeds the embedded features into a bilinear predictor to predict the synergy scores of drug combinations in different cancer cell lines. Specifically, LRGA network brings better model generalization ability, and effectively reduces the complexity of graph computation. The bilinear predictor facilitates the dimension transformation of the features and fuses the feature representation of the two drugs to improve the prediction performance. The loss function Smooth L1 effectively avoids gradient explosion, contributing to better model convergence. To validate the performance of DGSSynADR, we compare it with seven competitive methods. The comparison results demonstrate that DGSSynADR achieves better performance. Meanwhile, the prediction of DGSSynADR is validated by previous findings in case studies. Furthermore, detailed ablation studies indicate that the one-hot coding drug feature, LRGA model and bilinear predictor play a key role in improving the prediction performance.

AVAILABILITY AND IMPLEMENTATION

DGSSynADR is implemented in Python using the Pytorch machine-learning library, and it is freely available at https://github.com/DHUDBlab/DGSSynADR.

COIMMR: a computational framework to reveal the contribution of herbal ingredients against human cancer via immune microenvironment and metabolic reprogramming

Immune evasion and metabolism reprogramming have been regarded as two vital hallmarks of the mechanism of carcinogenesis. Thus, targeting the immune microenvironment and the reprogrammed metabolic processes will aid in developing novel anti-cancer drugs. In recent decades, herbal medicine has been widely utilized to treat cancer through the modulation of the immune microenvironment and reprogrammed metabolic processes. However, labor-based herbal ingredient screening is time consuming, laborious and costly. Luckily, some computational approaches have been proposed to screen candidates for drug discovery rapidly. Yet, it has been challenging to develop methods to screen drug candidates exclusively targeting specific pathways, especially for herbal ingredients which exert anti-cancer effects by multiple targets, multiple pathways and synergistic ways. Meanwhile, currently employed approaches cannot quantify the contribution of the specific pathway to the overall curative effect of herbal ingredients. Hence, to address this problem, this study proposes a new computational framework to infer the contribution of the immune microenvironment and metabolic reprogramming (COIMMR) in herbal ingredients against human cancer and specifically screen herbal ingredients targeting the immune microenvironment and metabolic reprogramming. Finally, COIMMR was applied to identify isoliquiritigenin that specifically regulates the T cells in stomach adenocarcinoma and cephaelin hydrochloride that specifically targets metabolic reprogramming in low-grade glioma. The in silico results were further verified using in vitro experiments. Taken together, our approach opens new possibilities for repositioning drugs targeting immune and metabolic dysfunction in human cancer and provides new insights for drug development in other diseases. COIMMR is available at https://github.com/LYN2323/COIMMR.

Staying Ahead of the Game: How SARS-CoV-2 has Accelerated the Application of Machine Learning in Pandemic Management

In recent years, machine learning (ML) techniques have garnered considerable interest for their potential use in accelerating the rate of drug discovery. With the emergence of the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) pandemic, the utilization of ML has become even more crucial in the search for effective antiviral medications. The pandemic has presented the scientific community with a unique challenge, and the rapid identification of potential treatments has become an urgent priority. Researchers have been able to accelerate the process of identifying drug candidates, repurposing existing drugs, and designing new compounds with desirable properties using machine learning in drug discovery. To train predictive models, ML techniques in drug discovery rely on the analysis of large datasets, including both experimental and clinical data. These models can be used to predict the biological activities, potential side effects, and interactions with specific target proteins of drug candidates. This strategy has proven to be an effective method for identifying potential coronavirus disease 2019 (COVID-19) and other disease treatments. This paper offers a thorough analysis of the various ML techniques implemented to combat COVID-19, including supervised and unsupervised learning, deep learning, and natural language processing. The paper discusses the impact of these techniques on pandemic drug development, including the identification of potential treatments, the understanding of the disease mechanism, and the creation of effective and safe therapeutics. The lessons learned can be applied to future outbreaks and drug discovery initiatives.

MultiGML: Multimodal graph machine learning for prediction of adverse drug events

Adverse drug events constitute a major challenge for the success of clinical trials. Several computational strategies have been suggested to estimate the risk of adverse drug events in preclinical drug development. While these approaches have demonstrated high utility in practice, they are at the same time limited to specific information sources. Thus, many current computational approaches neglect a wealth of information which results from the integration of different data sources, such as biological protein function, gene expression, chemical compound structure, cell-based imaging and others. In this work we propose an integrative and explainable multi-modal Graph Machine Learning approach (MultiGML), which fuses knowledge graphs with multiple further data modalities to predict drug related adverse events and general drug target-phenotype associations. MultiGML demonstrates excellent prediction performance compared to alternative algorithms, including various traditional knowledge graph embedding techniques. MultiGML distinguishes itself from alternative techniques by providing in-depth explanations of model predictions, which point towards biological mechanisms associated with predictions of an adverse drug event. Hence, MultiGML could be a versatile tool to support decision making in preclinical drug development.

Regulating the cell shift of endothelial cell-like myofibroblasts in pulmonary fibrosis

Pulmonary fibrosis is a common and severe fibrotic lung disease with high morbidity and mortality. Recent studies have reported a large number of unwanted myofibroblasts appearing in pulmonary fibrosis, and shown that the sustained activation of myofibroblasts is essential for unremitting interstitial fibrogenesis. However, the origin of these myofibroblasts remains poorly understood. Here, we create new mouse models of pulmonary fibrosis and identify a previously unknown population of endothelial cell (EC)-like myofibroblasts in normal lung tissue. We show that these EC-like myofibroblasts significantly contribute myofibroblasts to pulmonary fibrosis, which is confirmed by single-cell RNA sequencing of human pulmonary fibrosis. Using the transcriptional profiles, we identified a small molecule that redirects the differentiation of EC-like myofibroblasts and reduces pulmonary fibrosis in our mouse models. Our study reveals the mechanistic underpinnings of the differentiation of EC-like myofibroblasts in pulmonary fibrosis and may provide new strategies for therapeutic interventions.

CD16+ monocytes are involved in the hyper-inflammatory state of Prader-Willi Syndrome by single-cell transcriptomic analysis

Background

Patients with Prader-Willi syndrome (PWS) have a reduced life expectancy due to inflammation-related disease including cardiovascular disease and diabetes. Abnormal activation of peripheral immune system is postulated as a contributor. However, detailed features of the peripheral immune cells in PWS have not been fully elucidated.

Methods

Serum inflammatory cytokines were measured in healthy controls (n=13) and PWS patients (n=10) using a 65- multiplex cytokine assays. Changes of the peripheral immune cells in PWS was assessed by single-cell RNA sequencing (scRNA-seq) and high-dimensional mass cytometry (CyTOF) using peripheral blood mononuclear cells (PBMCs) from PWS patients (n=6) and healthy controls (n=12).

Results

PWS patients exhibited hyper-inflammatory signatures in PBMCs and monocytes were the most pronounced. Most inflammatory serum cytokines were increased in PWS, including IL-1β, IL-2R, IL-12p70, and TNF-α. The characteristics of monocytes evaluated by scRNA-seq and CyTOF showed that CD16⁺ monocytes were significantly increased in PWS patients. Functional pathway analysis revealed that CD16⁺ monocytes upregulated pathways in PWS were closely associated with TNF/IL-1β- driven inflammation signaling. The CellChat analysis identified CD16⁺ monocytes transmitted chemokine and cytokine signaling to drive inflammatory process in other cell types. Finally, we explored the PWS deletion region 15q11-q13 might be responsible for elevated levels of inflammation in the peripheral immune system.

Conclusion

The study highlights that CD16⁺ monocytes contributor to the hyper-inflammatory state of PWS which provides potential targets for immunotherapy in the future and expands our knowledge of peripheral immune cells in PWS at the single cell level for the first time.

Factor-specific generative pattern from large-scale drug-induced gene expression profile

Drug discovery is a complex and interdisciplinary field that requires the identification of potential drug targets for specific diseases. In this study, we present FacPat, a novel approach that identifies the optimal factor-specific pattern explaining the drug-induced gene expression profile. FacPat uses a genetic algorithm based on pattern distance to mine the optimal factor-specific pattern for each gene in the LINCS L1000 dataset. We applied Benjamini-Hochberg correction to control the false discovery rate and identified significant and interpretable factor-specific patterns consisting of 480 genes, 7 chemical compounds, and 38 human cell lines. Using our approach, we identified genes that show context-specific effects related to chemical compounds and/or human cell lines. Furthermore, we performed functional enrichment analysis to characterize biological features. We demonstrate that FacPat can be used to reveal novel relationships among drugs, diseases, and genes.

Transcriptome profile analysis reveals putative molecular mechanisms of 5-aminolevulinic acid toxicity

5-aminolevulinic acid (5-ALA) is the first precursor of the heme biosynthesis pathway, accumulated in acute intermittent porphyria (AIP), an inherited metabolic disease characterized by porphobilinogen deaminase deficiency. An increased incidence of hepatocellular carcinoma (HCC) has been reported as a long-term manifestation in symptomatic AIP patients. 5-ALA is an α-aminoketone prone to oxidation, yielding reactive oxygen species and 4,5-dioxovaleric acid. A high concentration of 5-ALA presents deleterious pro-oxidant potential. It can induce apoptosis, DNA damage, mitochondrial dysfunction, and altered expression of carcinogenesis-related proteins. Several hypotheses of the increased risk of HCC rely on the harmful effect of elevated 5-ALA in the liver of AIP patients, which could promote a pro-carcinogenic environment. We investigated the global transcriptional changes and perturbed molecular pathways in HepG2 cells following exposure to 5-ALA 25 mM for 2 h and 24 h using DNA microarray. Distinct transcriptome profiles were observed. 5-ALA '25 mM-2h' upregulated 10 genes associated with oxidative stress response and carcinogenesis. Enrichment analysis of differentially expressed genes by KEGG, Reactome, MetaCore™, and Gene Ontology, showed that 5-ALA '25 mM-24h' enriched pathways involved in drug detoxification, oxidative stress, DNA damage, cell death/survival, cell cycle, and mitochondria dysfunction corroborating the pro-oxidant properties of 5-ALA. Furthermore, our results disclosed other possible processes such as senescence, immune responses, endoplasmic reticulum stress, and also some putative effectors, such as sequestosome, osteopontin, and lon peptidase 1. This study provided additional knowledge about molecular mechanisms of 5-ALA toxicity which is essential to a deeper understanding of AIP and HCC pathophysiology. Furthermore, our findings can contribute to improving the efficacy of current therapies and the development of novel biomarkers and targets for diagnosis, prognosis, and therapeutic strategies for AHP/AIP and associated HCC.

Combination of furosemide, gold, and dopamine as a potential therapy for breast cancer

Breast cancer is one of the leading causes of death in women worldwide. Initially, it develops in the epithelium of the ducts or lobules of the breast glandular tissues with limited growth and the potential to metastasize. It is a highly heterogeneous malignancy; however, the common molecular mechanisms could help identify new targeted drugs for treating its subtypes. This study uses computational drug repositioning approaches to explore fresh drug candidates for breast cancer treatment. We also implemented reversal gene expression and gene expression-based signatures to explore novel drug candidates computationally. The drug activity profiles and related gene expression changes were acquired from the DrugBank, PubChem, and LINCS databases, and then in silico drug screening, molecular dynamics (MD) simulation, replica exchange MD simulations, and simulated annealing molecular dynamics (SAMD) simulations were conducted to discover and verify the valid drug candidates. We have found that compounds like furosemide, gold, and dopamine showed significant outcomes. Furthermore, the expression of genes related to breast cancer was observed to be reversed by these shortlisted drugs. Therefore, we postulate that combining furosemide, gold, and dopamine would be a potential combination therapy measurement for breast cancer patients.

Multivariate genome-wide association meta-analysis of over 1 million subjects identifies loci underlying multiple substance use disorders

Genetic liability to substance use disorders can be parsed into loci that confer general or substance-specific addiction risk. We report a multivariate genome-wide association meta-analysis that disaggregates general and substance-specific loci for published summary statistics of problematic alcohol use, problematic tobacco use, cannabis use disorder, and opioid use disorder in a sample of 1,025,550 individuals of European descent and 92,630 individuals of African descent. Nineteen independent SNPs were genome-wide significant (P < 5e-8) for the general addiction risk factor (addiction-rf), which showed high polygenicity. Across ancestries, PDE4B was significant (among other genes), suggesting dopamine regulation as a cross-substance vulnerability. An addiction-rf polygenic risk score was associated with substance use disorders, psychopathologies, somatic conditions, and environments associated with the onset of addictions. Substance-specific loci (9 for alcohol, 32 for tobacco, 5 for cannabis, 1 for opioids) included metabolic and receptor genes. These findings provide insight into genetic risk loci for substance use disorders that could be leveraged as treatment targets.

Computational drug repurposing by exploiting large-scale gene expression data: Strategy, methods and applications

De novo drug development is an extremely complex, time-consuming and costly task. Urgent needs for therapies of various diseases have greatly accelerated searches for more effective drug development methods. Luckily, drug repurposing provides a new and effective perspective on disease treatment. Rapidly increased large-scale transcriptome data paints a detailed prospect of gene expression during disease onset and thus has received wide attention in the field of computational drug repurposing. However, how to efficiently mine transcriptome data and identify new indications for old drugs remains a critical challenge. This review discussed the irreplaceable role of transcriptome data in computational drug repurposing and summarized some representative databases, tools and strategies. More importantly, it proposed a practical guideline through establishing the correspondence between three gene expression data types and five strategies, which would facilitate researchers to adopt appropriate strategies to deeply mine large-scale transcriptome data and discover more effective therapies.

Impact of somatic mutations and transcriptomic alterations on cancer aneuploidy

Aneuploidy has been recognized as one of hallmark of tumorigenesis since the early 20th century. Recent developments in structural variation analysis in the human genome have revealed the diversity of aneuploidy in cancer. However, the effects of gene mutation and expression in tumors on aneuploidy remain poorly understood. Here, we performed whole exome analysis of over 5,000 Japanese cancer cases and investigated the impact of somatic mutations and gene expression alterations on aneuploidy. First, we evaluated tumor content and genomic alterations that could influence aneuploidy. Next, we compared the aneuploidy frequency in 18 cancer types and observed that TP53 mutations were associated with the aneuploidy on specific chromosomes in colorectal and gastric cancers. Finally, we used expression analysis to isolate pathways involved in aneuploidy accumulation from tumors without TP53 mutations. Chromosomal instability and cell cycle aberration were associated with aneuploidy in TP53 wild-type tumors, and 26 commonly upregulated genes were identified in aneuploidy-high solid tumors without TP53 mutations. Among them, two cancer-related genes (CENPA and PBK) were involved in aneuploidy. Our integrated analysis revealed that both TP53 mutations and transcriptomic alterations independent of somatic mutations affect aneuploidy accumulation. Our findings will facilitate further understanding of diverse aneuploidies in the tumorigenesis.

An explorative study for leveraging transcriptomic data of embryonic stem cells in mining cancer stemness genes, regulators, and networks

Due to the exquisite ability of cancer stemness to facilitate tumor initiation, metastasis, and cancer therapy resistance, targeting cancer stemness is expected to have clinical implications for cancer treatment. Genes are fundamental for forming and maintaining stemness. Considering shared genetic programs and pathways between embryonic stem cells and cancer stem cells, we conducted a study analyzing transcriptomic data of embryonic stem cells for mining potential cancer stemness genes. Firstly, we integrated co-expression and regression models and predicted 820 stemness genes. Results of gene enrichment analysis confirmed the good prediction performance for enriched signatures in cancer stem cells. Secondly, we provided an application case using the predicted stemness genes to construct a breast cancer stemness network. Mining on the network identified CD44, SOX2, TWIST1, and DLG4 as potential regulators of breast cancer stemness. Thirdly, using the signature of 31,028 chemical perturbations and their correlation with stemness marker genes, we predicted 67 stemness inhibitors with reasonable accuracy of 78%. Two drugs, namely Rigosertib and Proscillaridin A, were first identified as potential stemness inhibitors for melanoma and colon cancer, respectively. Overall, mining embryonic stem cell data provides a valuable way to identify cancer stemness regulators.

metaLINCS: an R package for meta-level analysis of LINCS L1000 drug signatures using stratified connectivity mapping

Summary

Accessing the collection of perturbed gene expression profiles, such as the LINCS L1000 connectivity map, is usually performed at the individual dataset level, followed by a summary performed by counting individual hits for each perturbagen. With the metaLINCS R package, we present an alternative approach that combines rank correlation and gene set enrichment analysis to identify meta-level enrichment at the perturbagen level and, in the case of drugs, at the mechanism of action level. This significantly simplifies the interpretation and highlights overarching themes in the data. We demonstrate the functionality of the package and compare its performance against those of three currently used approaches.

Availability and implementation

metaLINCS is released under GPL3 license. Source code and documentation are freely available on GitHub (https://github.com/bigomics/metaLINCS).

Supplementary information

Supplementary data are available at Bioinformatics Advances online.

Biomedical Data and Deep Learning Computational Models for Predicting Compound-Protein Relations

The identification of compound-protein relations (CPRs), which includes compound-protein interactions (CPIs) and compound-protein affinities (CPAs), is critical to drug development. A common method for compound-protein relation identification is the use of in vitro screening experiments. However, the number of compounds and proteins is massive, and in vitro screening experiments are labor-intensive, expensive, and time-consuming with high failure rates. Researchers have developed a computational field called virtual screening (VS) to aid experimental drug development. These methods utilize experimentally validated biological interaction information to generate datasets and use the physicochemical and structural properties of compounds and target proteins as input information to train computational prediction models. At present, deep learning has been widely used in computer vision and natural language processing and has experienced epoch-making progress. At the same time, deep learning has also been used in the field of biomedicine widely, and the prediction of CPRs based on deep learning has developed rapidly and has achieved good results. The purpose of this study is to investigate and discuss the latest applications of deep learning techniques in CPR prediction. First, we describe the datasets and feature engineering (i.e., compound and protein representations and descriptors) commonly used in CPR prediction methods. Then, we review and classify recent deep learning approaches in CPR prediction. Next, a comprehensive comparison is performed to demonstrate the prediction performance of representative methods on classical datasets. Finally, we discuss the current state of the field, including the existing challenges and our proposed future directions. We believe that this investigation will provide sufficient references and insight for researchers to understand and develop new deep learning methods to enhance CPR predictions.

From drug repositioning to target repositioning: prediction of therapeutic targets using genetically perturbed transcriptomic signatures

Motivation

A critical element of drug development is the identification of therapeutic targets for diseases. However, the depletion of therapeutic targets is a serious problem.

Results

In this study, we propose the novel concept of target repositioning, an extension of the concept of drug repositioning, to predict new therapeutic targets for various diseases. Predictions were performed by a trans-disease analysis which integrated genetically perturbed transcriptomic signatures (knockdown of 4345 genes and overexpression of 3114 genes) and disease-specific gene transcriptomic signatures of 79 diseases. The trans-disease method, which takes into account similarities among diseases, enabled us to distinguish the inhibitory from activatory targets and to predict the therapeutic targetability of not only proteins with known target–disease associations but also orphan proteins without known associations. Our proposed method is expected to be useful for understanding the commonality of mechanisms among diseases and for therapeutic target identification in drug discovery.

Availability and implementation

Supplemental information and software are available at the following website [http://labo.bio.kyutech.ac.jp/~yamani/target_repositioning/].

Supplementary information

Supplementary data are available at Bioinformatics online.

Vitamin D Receptor Activation Attenuates Hippo Pathway Effectors and Cell Survival in Metastatic Neuroblastoma

Survival for high-risk neuroblastoma remains poor. Most patients who recur, present with metastatic disease, and few targetable pathways that govern spread to distant sites are currently known. We previously developed a metastatic mouse model to select cells with enhanced ability to spread to the bone and brain and identified a signature based on differentially expressed genes, which also predicted patient survival. To discover new neuroblastoma therapies, we utilized the Connectivity Map to identify compounds that can reverse this metastatic transcriptional signature and found calcipotriol, a vitamin D3 analog, to be a compound that selectively targets cell lines with enhanced metastatic potential. Calcipotriol treatment of enhanced metastatic, but not parental, cells reduces proliferation and survival via vitamin D receptor (VDR) signaling, increases the expression of RASSF2, a negative regulator of the Hippo signaling pathway, and reduces the levels of the Hippo pathway effectors YAP and TAZ. RASSF2 is required for the effects of calcipotriol and for the reduction of levels and nuclear localization of YAP/TAZ. Migration of the enhanced metastatic cells and YAP/TAZ levels are reduced after calcipotriol treatment and YAP overexpression reduces calcipotriol sensitivity. Furthermore, metastatic cells that overexpress VDR also showed lower tumor burden in vivo.

IMPLICATIONS

This newly identified link between VDR signaling and the Hippo pathway could inform treatment strategies for metastatic neuroblastoma.

Combining phenotypic profiling and targeted RNA-Seq reveals linkages between transcriptional perturbations and chemical effects on cell morphology: Retinoic acid as an example

The United States Environmental Protection Agency has proposed a tiered testing strategy for chemical hazard evaluation based on new approach methods (NAMs). The first tier includes in vitro profiling assays applicable to many (human) cell types, such as high-throughput transcriptomics (HTTr) and high-throughput phenotypic profiling (HTPP). The goals of this study were to: (1) harmonize the seeding density of U-2 OS human osteosarcoma cells for use in both assays; (2) compare HTTr- versus HTPP-derived potency estimates for 11 mechanistically diverse chemicals; (3) identify candidate reference chemicals for monitoring assay performance in future screens; and (4) characterize the transcriptional and phenotypic changes in detail for all-trans retinoic acid (ATRA) as a model compound known for its adverse effects on osteoblast differentiation. The results of this evaluation showed that (1) HTPP conducted at low (400 cells/well) and high (3000 cells/well) seeding densities yielded comparable potency estimates and similar phenotypic profiles for the tested chemicals; (2) HTPP and HTTr resulted in comparable potency estimates for changes in cellular morphology and gene expression, respectively; (3) three test chemicals (etoposide, ATRA, dexamethasone) produced concentration-dependent effects on cellular morphology and gene expression that were consistent with known modes-of-action, demonstrating their suitability for use as reference chemicals for monitoring assay performance; and (4) ATRA produced phenotypic changes that were highly similar to other retinoic acid receptor activators (AM580, arotinoid acid) and some retinoid X receptor activators (bexarotene, methoprene acid). This phenotype was observed concurrently with autoregulation of the RARB gene. Both effects were prevented by pre-treating U-2 OS cells with pharmacological antagonists of their respective receptors. Thus, the observed phenotype could be considered characteristic of retinoic acid pathway activation in U-2 OS cells. These findings lay the groundwork for combinatorial screening of chemicals using HTTr and HTPP to generate complementary information for the first tier of a NAM-based chemical hazard evaluation strategy.

Single-molecule long-read sequencing analysis improves genome annotation and sheds new light on the transcripts and splice isoforms of Zoysia japonica

BACKGROUND

Zoysia japonica is an important warm-season turfgrass used worldwide. Although the draft genome sequence and a vast amount of next-generation sequencing data have been published, the current genome annotation and complete mRNA structure remain incomplete. Therefore, to analyze the full-length transcriptome of Z. japonica, we used the PacBio single-molecule long-read sequencing method in this study.

RESULTS

First, we generated 37,056 high-confidence non-redundant transcripts from 16,005 gene loci. Next, 32,948 novel transcripts, 913 novel gene loci, 8035 transcription factors, 89 long non-coding RNAs, and 254 fusion transcripts were identified. Furthermore, 15,675 alternative splicing events and 5325 alternative polyadenylation sites were detected. In addition, using bioinformatics analysis, the underlying transcriptional mechanism of senescence was explored based on the revised reference transcriptome.

CONCLUSION

This study provides a full-length reference transcriptome of Z. japonica using PacBio single-molecule long-read sequencing for the first time. These results contribute to our knowledge of the transcriptome and improve the knowledge of the reference genome of Z. japonica. This will also facilitate genetic engineering projects using Z. japonica.

Probing the Immune System Dynamics of the COVID-19 Disease for Vaccine Designing and Drug Repurposing Using Bioinformatics Tools

The pathogenesis of COVID-19 is complicated by immune dysfunction. The impact of immune-based therapy in COVID-19 patients has been well documented, with some notable studies on the use of anti-cytokine medicines. However, the complexity of disease phenotypes, patient heterogeneity and the varying quality of evidence from immunotherapy studies provide problems in clinical decision-making. This review seeks to aid therapeutic decision-making by giving an overview of the immunological responses against COVID-19 disease that may contribute to the severity of the disease. We have extensively discussed theranostic methods for COVID-19 detection. With advancements in technology, bioinformatics has taken studies to a higher level. The paper also discusses the application of bioinformatics and machine learning tools for the diagnosis, vaccine design and drug repurposing against SARS-CoV-2.

Combating Viral Diseases in the Era of Systems Medicine

Viruses can cause many diseases resulting in disabilities and death. Fortunately, advances in systems medicine enable the development of effective therapies for treating viral diseases, of vaccines to prevent viral infections, as well as of diagnostic tools to mitigate the risk and reduce the death toll. Characterizing the SARS-CoV-2 gene sequence and the role of its spike protein in infection informs development of small molecule drugs, antibodies, and vaccines to combat infection and complication, as well as to end the pandemic. Drug repurposing of small molecule drugs is a viable strategy to combat viral diseases; the key concepts include (1) linking a drug candidate's pharmacological network to its pharmacodynamic response in patients; (2) linking a drug candidate's physicochemical properties to its pharmacokinetic characteristics; and (3) optimizing the safe and effective dosing regimen within its therapeutic window. Computational integration of drug-induced signaling pathways with clinical outcomes is useful to inform selection of potential drug candidates with respect to safety and effectiveness. Key pharmacokinetic and pharmacodynamic principles for computational optimization of drug development include a drug candidate's Cminss/IC95 ratio, pharmacokinetic characteristics, and systemic exposure-response relationship, where Cminss is the trough concentration following multiple dosing. In summary, systems medicine approaches play a vital role in global success in combating viral diseases, including global real-time information sharing, development of test kits, drug repurposing, discovery and development of safe, effective therapies, detection of highly transmissible and deadly variants, and development of vaccines.

Past, Present, and Future of Therapies for Pituitary Neuroendocrine Tumors: Need for Omics and Drug Repositioning Guidance

Innovation roadmaps are important, because they encourage the actors in an innovation ecosystem to creatively imagine multiple possible science future(s), while anticipating the prospects and challenges on the innovation trajectory. In this overarching context, this expert review highlights the present unmet need for therapeutic innovations for pituitary neuroendocrine tumors (PitNETs), also known as pituitary adenomas. Although there are many drugs used in practice to treat PitNETs, many of these drugs can have negative side effects and show highly variable outcomes in terms of overall recovery. Building innovation roadmaps for PitNETs' treatments can allow incorporation of systems biology approaches to bring about insights at multiple levels of cell biology, from genes to proteins to metabolites. Using the systems biology techniques, it will then be possible to offer potential therapeutic strategies for the convergence of preventive approaches and patient-centered disease treatment. Here, we first provide a comprehensive overview of the molecular subtypes of PitNETs and therapeutics for these tumors from the past to the present. We then discuss examples of clinical trials and drug repositioning studies and how multi-omics studies can help in discovery and rational development of new therapeutics for PitNETs. Finally, this expert review offers new public health and personalized medicine approaches on cases that are refractory to conventional treatment or recur despite currently used surgical and/or drug therapy.

Hepatocellular carcinoma after a sustained virological response by direct-acting antivirals harbors TP53 inactivation

Introduction

The genomic characteristics of hepatocellular carcinoma (HCC) after a sustained virological response (SVR) and its differences according to whether an SVR was achieved by treatment with direct‐acting antivirals (DAA) or interferon (IFN) are still not fully understood.

Methods

Sixty‐nine surgically resected HCCs from patients with hepatitis C virus infection were analyzed by gene expression profiling and whole‐exome sequencing.

Results

Among the 69 HCC patients, 34 HCCs in which an SVR was not achieved at the time of surgery were classified as HCV‐positive, and 35 HCCs in which an SVR was achieved at the time of surgery were classified as HCV‐SVR. According to the HCV treatment, 35 HCV‐SVR HCCs were classified into two groups: eight tumors with DAA (HCV‐SVR‐DAA) and 24 tumors with interferon (HCV‐SVR‐IFN). The frequency of samples with ARID2 mutations was significantly lower in HCV‐SVR than in HCV‐positive tumors (p = 0.048). In contrast, the frequency of samples with PREX2 mutations was significantly higher in HCV‐SVR samples than in HCV‐positive samples (p = 0.048). Among the patients with HCV‐SVR, the frequency of samples with TP53 mutations was significantly higher in HCV‐SVR‐DAA tumors than in HCV‐SVR‐IFN tumors (p = 0.030). TP53 inactivation scores in HCV‐SVR‐DAA tumors were found to be significantly enhanced in comparison to HCV‐SVR‐IFN tumors (p = 0.022). In addition, chromosomal instability and PI3K/AKT/mTOR pathway signatures were enhanced in HCV‐SVR‐DAA tumors. HCV‐SVR‐DAA was significantly associated with portal vein invasion (p = 0.003) in comparison to HCV‐SVR‐IFN.

Conclusion

Our dataset potentially serves as a fundamental resource for the genomic characteristics of HCV‐SVR‐DAA tumors. Our comprehensive genetic profiling by WES revealed significant differences in the mutation rate of several driver genes between HCV‐positive tumors and HCV‐SVR tumors. Furthermore, it was revealed that the frequency of samples with mutations in TP53 was significantly higher in HCV‐SVR‐DAA tumors than in HCV‐SVR‐IFN tumors.

An Integrated In Silico, In Vitro and Tumor Tissues Study Identified Selenoprotein S (SELENOS) and Valosin-Containing Protein (VCP/p97) as Novel Potential Associated Prognostic Biomarkers in Triple Negative Breast Cancer

Simple Summary

Triple negative breast cancer (TNBC) represents a clinical challenge because its early relapse, poor overall survival and lack of effective treatments. Altered levels selenoproteins have been correlated with development and progression of some cancers, however, no consistent data are available about their involvement in TNBC. Here we analyzed the expression profile of all twenty-five human selenoproteins in TNBC cells and tissues by a systematic approach, integrating in silico and wet lab approaches. We showed that the expression profiles of five selenoproteins are specifically dysregulated in TNBC. Most importantly, by a bioinformatics analysis, we selected selenoprotein S and its interacting protein valosin-containing protein (VCP/p97) as inter-related with the others and whose coordinated over-expression is associated with poor prognosis in TNBC. Overall, we highlighted two mechanistically related novel proteins whose correlated expression could be exploited for a better definition of prognosis as well as suggested as novel therapeutic target in TNBC.

Abstract

Background. Triple negative breast cancer (TNBC) is a heterogeneous group of tumors with early relapse, poor overall survival, and lack of effective treatments. Hence, new prognostic biomarkers and therapeutic targets are needed. Methods. The expression profile of all twenty-five human selenoproteins was analyzed in TNBC by a systematic approach.In silicoanalysis was performed on publicly available mRNA expression datasets (Cancer Cell Line Encyclopedia, CCLE and Library of Integrated Network-based Cellular Signatures, LINCS). Reverse transcription quantitative PCR analysis evaluated selenoprotein mRNA expression in TNBC versus non-TNBC and normal breast cells, and in TNBC tissues versus normal counterparts. Immunohistochemistry was employed to study selenoproteins in TNBC tissues. STRING and Cytoscape tools were used for functional and network analysis. Results.GPX1, GPX4, SELENOS, TXNRD1 and TXNRD3 were specifically overexpressed in TNBC cells, tissues and CCLE/LINCS datasets. Network analysis demonstrated that SELENOS-binding valosin-containing protein (VCP/p97) played a critical hub role in the TNBCselenoproteins sub-network, being directly associated with SELENOS expression. The combined overexpression of SELENOS and VCP/p97 correlated with advanced stages and poor prognosis in TNBC tissues and the TCGA dataset. Conclusion. Combined evaluation of SELENOS and VCP/p97 might represent a novel potential prognostic signature and a therapeutic target to be exploited in TNBC.

Epigenetic landscape of drug responses revealed through large-scale ChIP-seq data analyses

Background

Elucidating the modes of action (MoAs) of drugs and drug candidate compounds is critical for guiding translation from drug discovery to clinical application. Despite the development of several data-driven approaches for predicting chemical–disease associations, the molecular cues that organize the epigenetic landscape of drug responses remain poorly understood.

Results

With the use of a computational method, we attempted to elucidate the epigenetic landscape of drug responses, in terms of transcription factors (TFs), through large-scale ChIP-seq data analyses. In the algorithm, we systematically identified TFs that regulate the expression of chemically induced genes by integrating transcriptome data from chemical induction experiments and almost all publicly available ChIP-seq data (consisting of 13,558 experiments). By relating the resultant chemical–TF associations to a repository of associated proteins for a wide range of diseases, we made a comprehensive prediction of chemical–TF–disease associations, which could then be used to account for drug MoAs. Using this approach, we predicted that: (1) cisplatin promotes the anti-tumor activity of TP53 family members but suppresses the cancer-inducing function of MYCs; (2) inhibition of RELA and E2F1 is pivotal for leflunomide to exhibit antiproliferative activity; and (3) CHD8 mediates valproic acid-induced autism.

Conclusions

Our proposed approach has the potential to elucidate the MoAs for both approved drugs and candidate compounds from an epigenetic perspective, thereby revealing new therapeutic targets, and to guide the discovery of unexpected therapeutic effects, side effects, and novel targets and actions.

A Hybrid Protocol for Identifying Comorbidity-Based Potential Drugs for COVID-19 Using Biomedical Literature Mining, Network Analysis, and Deep Learning

Coronavirus disease 2019 (COVID-19) caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV2) has spread on an unprecedented scale around the globe. Despite of 141,975 published papers on COVID-19 and several hundreds of new studies carried out every day, this pandemic remains as a global challenge. Biomedical literature mining helps the researchers to understand the etiology of the disease and to gain an in-depth knowledge of the disease, potential drugs, vaccines developed and novel therapies. In addition to the available treatments, there is a huge need to address the comorbidity-based disease mortality in case of COVID-19 patients with type 2 diabetes mellitus (T2D), hypertension and cardiovascular disease (CVD). In this chapter, we provide a hybrid protocol based on biomedical literature mining, network analysis of omics data, and deep learning for the identification of most potential drugs for COVID-19.

High-Throughput Strategies for the Discovery of Anticancer Drugs by Targeting Transcriptional Reprogramming

Transcriptional reprogramming contributes to the progression and recurrence of cancer. However, the poorly elucidated mechanisms of transcriptional reprogramming in tumors make the development of effective drugs difficult, and gene expression signature is helpful for connecting genetic information and pharmacologic treatment. So far, there are two gene-expression signature-based high-throughput drug discovery approaches: L1000, which measures the mRNA transcript abundance of 978 "landmark" genes, and high-throughput sequencing-based high-throughput screening (HTS²); they are suitable for anticancer drug discovery by targeting transcriptional reprogramming. L1000 uses ligation-mediated amplification and hybridization to Luminex beads and highlights gene expression changes by detecting bead colors and fluorescence intensity of phycoerythrin signal. HTS² takes advantage of RNA-mediated oligonucleotide annealing, selection, and ligation, high throughput sequencing, to quantify gene expression changes by directly measuring gene sequences. This article summarizes technological principles and applications of L1000 and HTS², and discusses their advantages and limitations in anticancer drug discovery.

Improvement of the performance of anticancer peptides using a drug repositioning pipeline

The use of anticancer peptides (ACPs) as an alternative/complementary strategy to conventional chemotherapy treatments has been shown to decrease drug resistance and/or severe side effects. However, the efficacy of the positively-charged ACP is inhibited by elevated levels of negatively-charged cell-surface components which trap the peptides and prevent their contact with the cell membrane. Consequently, this decreases ACP-mediated membrane pore formation and cell lysis. Negatively-charged heparan sulphate (HS) and chondroitin sulphate (CS) have been shown to inhibit the cytotoxic effect of ACPs. In this study, we propose a strategy to promote the broad utilization of ACPs. In this context, we developed a drug repositioning pipeline to analyse transcriptomics data generated for four different cancer cell lines (A549, HEPG2, HT29, and MCF7) treated with hundreds of drugs in the LINCS L1000 project. Based on previous studies identifying genes modulating levels of the glycosaminoglycans (GAGs) HS and CS at the cell surface, our analysis aimed at identifying drugs inhibiting genes correlated with high HS and CS levels. As a result, we identified six chemicals as likely repositionable drugs with the potential to enhance the performance of ACPs. The codes in R and Python programming languages are publicly available in https://github.com/ElyasMo/ACPs_HS_HSPGs_CS. As a conclusion, these six drugs are highlighted as excellent targets for synergistic studies with ACPs aimed at lowering the costs associated with ACP-treatment.

TRIOMPHE: Transcriptome-Based Inference and Generation of Molecules with Desired Phenotypes by Machine Learning

One of the most challenging tasks in the drug-discovery process is the efficient identification of small molecules with desired phenotypes. In this study, we propose a novel computational method for omics-based de novo drug design, which we call TRIOMPHE (transcriptome-based inference and generation of molecules with desired phenotypes). We investigated the correlation between chemically induced transcriptome profiles (reflecting cellular responses to compound treatment) and genetically perturbed transcriptome profiles (reflecting cellular responses to gene knock-down or gene overexpression of target proteins) in terms of ligand-target interactions. Subsequently, we developed novel machine learning methods to generate the chemical structures of new molecules with desired transcriptome profiles in the framework of a variational autoencoder. The use of desired transcriptome profiles enables the automatic design of molecules that are likely to have bioactivities for target proteins of interest. We showed that our methods can generate chemically valid molecules that are likely to have biological activities on 10 target proteins; moreover, they can outperform previous methods that had the same objective. Our omics-based structure generator is expected to be useful for the de novo design of drugs for a variety of target proteins.

Deep fusion learning facilitates anatomical therapeutic chemical recognition in drug repurposing and discovery

The advent of large-scale biomedical data and computational algorithms provides new opportunities for drug repurposing and discovery. It is of great interest to find an appropriate data representation and modeling method to facilitate these studies. The anatomical therapeutic chemical (ATC) classification system, proposed by the World Health Organization (WHO), is an essential source of information for drug repurposing and discovery. Besides, computational methods are applied to predict drug ATC classification. We conducted a systematic review of ATC computational prediction studies and revealed the differences in data sets, data representation, algorithm approaches, and evaluation metrics. We then proposed a deep fusion learning (DFL) framework to optimize the ATC prediction model, namely DeepATC. The methods based on graph convolutional network, inferring biological network and multimodel attentive fusion network were applied in DeepATC to extract the molecular topological information and low-dimensional representation from the molecular graph and heterogeneous biological networks. The results indicated that DeepATC achieved superior model performance with area under the curve (AUC) value at 0.968. Furthermore, the DFL framework was performed for the transcriptome data-based ATC prediction, as well as another independent task that is significantly relevant to drug discovery, namely drug-target interaction. The DFL-based model achieved excellent performance in the above-extended validation task, suggesting that the idea of aggregating the heterogeneous biological network and node's (molecule or protein) self-topological features will bring inspiration for broader drug repurposing and discovery research.

Discriminant Projection Shared Dictionary Learning for Classification of Tumors Using Gene Expression Data

With a variety of tumor subtypes, personalized treatments need to identify the subtype of a tumor as accurately as possible. The development of DNA microarrays provides an opportunity to predict tumor classification. One strategy is to use gene expression profiling to extend current biological insights into the disease. However, overfitting problems exist in most machine learning methods when classifying tumor gene expression profile data characterized by high dimensional, small samples and nonlinearities. As a new machine learning methods, dictionary learning has become a more effective algorithm for gene expression profile classification. Here, a new method called discriminant projection shared dictionary learning (DPSDL) is proposed for classifying tumor subtypes using LINCS gene expression profile data. The method trains a shared dictionary, embeds Fisher discriminant criteria to obtain a class-specific sub-dictionary and coding coefficients. At the same time, a projection matrix is trained to widen the distance between different classes of samples. Experimental results show that our method performs better classification based on gene expression profile than the other dictionary learning methods and machine learning methods.

In vitro model of perimenopausal depression implicates steroid metabolic and proinflammatory genes

The estimated 20-30% of women who develop perimenopausal depression (PMD) are at an increased risk of cardiovascular and all-cause mortality. The therapeutic benefits of estradiol (E2) and symptom-provoking effects of E2-withdrawal (E2-WD) suggest that a greater sensitivity to changes in E2 at the cellular level contribute to PMD. We developed an in vitro model of PMD with lymphoblastoid cell lines (LCLs) derived from participants of a prior E2-WD clinical study. LCLs from women with past PMD (n = 8) or control women (n = 9) were cultured in three experimental conditions: at vehicle baseline, during E2 treatment, and following E2-WD. Transcriptome analysis revealed significant differences in transcript expression in PMD in all experimental conditions, and significant overlap in genes that were changed in PMD regardless of experimental condition. Of these, chemokine CXCL10, previously linked to cardiovascular disease, was upregulated in women with PMD, but most so after E2-WD (p < 1.55 × 10^-5). CYP7B1, an enzyme intrinsic to DHEA metabolism, was upregulated in PMD across experimental conditions (F_(1,45) = 19.93, p < 0.0001). These transcripts were further validated via qRT-PCR. Gene networks dysregulated in PMD included inflammatory response, early/late E2-response, and cholesterol homeostasis. Our results provide evidence that differential behavioral responsivity to E2-WD in PMD reflects intrinsic differences in cellular gene expression. Genes such as CXCL10, CYP7B1, and corresponding proinflammatory and steroid biosynthetic gene networks, may represent biomarkers and molecular targets for intervention in PMD. Finally, this in vitro model allows for future investigations into the mechanisms of genes and gene networks involved in the vulnerability to, and consequences of, PMD.

Scoring functions for drug-effect similarity

MOTIVATION

The difficulty to find new drugs and bring them to the market has led to an increased interest to find new applications for known compounds. Biological samples from many disease contexts have been extensively profiled by transcriptomics, and, intuitively, this motivates to search for compounds with a reversing effect on the expression of characteristic disease genes. However, disease effects may be cell line-specific and also depend on other factors, such as genetics and environment. Transcription profile changes between healthy and diseased cells relate in complex ways to profile changes gathered from cell lines upon stimulation with a drug. Despite these differences, we expect that there will be some similarity in the gene regulatory networks at play in both situations. The challenge is to match transcriptomes for both diseases and drugs alike, even though the exact molecular pathology/pharmacogenomics may not be known.

RESULTS

We substitute the challenge to match a drug effect to a disease effect with the challenge to match a drug effect to the effect of the same drug at another concentration or in another cell line. This is welldefined, reproducible in vitro and in silico and extendable with external data. Based on the Connectivity Map (CMap) dataset, we combined 26 different similarity scores with six different heuristics to reduce the number of genes in the model. Such gene filters may also utilize external knowledge e.g. from biological networks. We found that no similarity score always outperforms all others for all drugs, but the Pearson correlation finds the same drug with the highest reliability. Results are improved by filtering for highly expressed genes and to a lesser degree for genes with large fold changes. Also a network-based reduction of contributing transcripts was beneficial, here implemented by the FocusHeuristics. We found no drop in prediction accuracy when reducing the whole transcriptome to the set of 1000 landmark genes of the CMap's successor project Library of Integrated Network-based Cellular Signatures. All source code to re-analyze and extend the CMap data, the source code of heuristics, filters and their evaluation are available to propel the development of new methods for drug repurposing.

AVAILABILITY

https://bitbucket.org/ibima/moldrugeffectsdb.

CONTACT

steffen.moeller@uni-rostock.de.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Briefings in Bioinformatics online.

Reconciling multiple connectivity scores for drug repurposing

The basis of several recent methods for drug repurposing is the key principle that an efficacious drug will reverse the disease molecular ‘signature’ with minimal side effects. This principle was defined and popularized by the influential ‘connectivity map’ study in 2006 regarding reversal relationships between disease- and drug-induced gene expression profiles, quantified by a disease-drug ‘connectivity score.’ Over the past 15 years, several studies have proposed variations in calculating connectivity scores toward improving accuracy and robustness in light of massive growth in reference drug profiles. However, these variations have been formulated inconsistently using various notations and terminologies even though they are based on a common set of conceptual and statistical ideas. Therefore, we present a systematic reconciliation of multiple disease-drug similarity metrics (\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{upgreek} \usepackage{mathrsfs} \setlength{\oddsidemargin}{-69pt} \begin{document} }{}$ES$\end{document}, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{upgreek} \usepackage{mathrsfs} \setlength{\oddsidemargin}{-69pt} \begin{document} }{}$css$\end{document}, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{upgreek} \usepackage{mathrsfs} \setlength{\oddsidemargin}{-69pt} \begin{document} }{}$Sum$\end{document}, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{upgreek} \usepackage{mathrsfs} \setlength{\oddsidemargin}{-69pt} \begin{document} }{}$Cosine$\end{document}, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{upgreek} \usepackage{mathrsfs} \setlength{\oddsidemargin}{-69pt} \begin{document} }{}$XSum$\end{document}, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{upgreek} \usepackage{mathrsfs} \setlength{\oddsidemargin}{-69pt} \begin{document} }{}$XCor$\end{document}, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{upgreek} \usepackage{mathrsfs} \setlength{\oddsidemargin}{-69pt} \begin{document} }{}$XSpe$\end{document}, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{upgreek} \usepackage{mathrsfs} \setlength{\oddsidemargin}{-69pt} \begin{document} }{}$XCos$\end{document}, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{upgreek} \usepackage{mathrsfs} \setlength{\oddsidemargin}{-69pt} \begin{document} }{}$EWCos$\end{document}) and connectivity scores (\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{upgreek} \usepackage{mathrsfs} \setlength{\oddsidemargin}{-69pt} \begin{document} }{}$CS$\end{document}, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{upgreek} \usepackage{mathrsfs} \setlength{\oddsidemargin}{-69pt} \begin{document} }{}$RGES$\end{document}, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{upgreek} \usepackage{mathrsfs} \setlength{\oddsidemargin}{-69pt} \begin{document} }{}$NCS$\end{document}, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{upgreek} \usepackage{mathrsfs} \setlength{\oddsidemargin}{-69pt} \begin{document} }{}$WCS$\end{document}, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{upgreek} \usepackage{mathrsfs} \setlength{\oddsidemargin}{-69pt} \begin{document} }{}$Tau$\end{document}, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{upgreek} \usepackage{mathrsfs} \setlength{\oddsidemargin}{-69pt} \begin{document} }{}$CSS$\end{document}, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{upgreek} \usepackage{mathrsfs} \setlength{\oddsidemargin}{-69pt} \begin{document} }{}$EMUDRA$\end{document}) by defining them using consistent notation and terminology. In addition to providing clarity and deeper insights, this coherent definition of connectivity scores and their relationships provides a unified scheme that newer methods can adopt, enabling the computational drug-development community to compare and investigate different approaches easily. To facilitate the continuous and transparent integration of newer methods, this article will be available as a live document (https://jravilab.github.io/connectivity_scores) coupled with a GitHub repository (https://github.com/jravilab/connectivity_scores) that any researcher can build on and push changes to.

Dynamic flux balance analysis of whole-body metabolism for type 1 diabetes

Type 1 diabetes (T1D) mellitus is a systemic disease triggered by a local autoimmune inflammatory reaction in insulin-producing cells that induce organ-wide, long-term metabolic effects. Mathematical modeling of the whole-body regulatory bihormonal system has helped to identify therapeutic interventions but is limited to a coarse-grained representation of metabolism. To extend the depiction of T1D, we developed a whole-body model of organ-specific regulation and metabolism that highlighted chronic inflammation as a hallmark of the disease, identified processes related to neurodegenerative disorders and suggested calcium channel blockers as adjuvants for diabetes control. In addition, whole-body modeling of a patient population allowed for the assessment of between-individual variability to insulin and suggested that peripheral glucose levels are degenerate biomarkers of the internal metabolic state. Taken together, the organ-resolved, dynamic modeling approach enables modeling and simulation of metabolic disease at greater levels of coverage and precision and the generation of hypothesis from a molecular level up to the population level.

Drug-Target Interaction Prediction Based on Adversarial Bayesian Personalized Ranking

The prediction of drug-target interaction (DTI) is a key step in drug repositioning. In recent years, many studies have tried to use matrix factorization to predict DTI, but they only use known DTIs and ignore the features of drug and target expression profiles, resulting in limited prediction performance. In this study, we propose a new DTI prediction model named AdvB-DTI. Within this model, the features of drug and target expression profiles are associated with Adversarial Bayesian Personalized Ranking through matrix factorization. Firstly, according to the known drug-target relationships, a set of ternary partial order relationships is generated. Next, these partial order relationships are used to train the latent factor matrix of drugs and targets using the Adversarial Bayesian Personalized Ranking method, and the matrix factorization is improved by the features of drug and target expression profiles. Finally, the scores of drug-target pairs are achieved by the inner product of latent factors, and the DTI prediction is performed based on the score ranking. The proposed model effectively takes advantage of the idea of learning to rank to overcome the problem of data sparsity, and perturbation factors are introduced to make the model more robust. Experimental results show that our model could achieve a better DTI prediction performance.

Oncoprotein-specific molecular interaction maps (SigMaps) for cancer network analyses

Tumor-specific elucidation of physical and functional oncoprotein interactions could improve tumorigenic mechanism characterization and therapeutic response prediction. Current interaction models and pathways, however, lack context specificity and are not oncoprotein specific. We introduce SigMaps as context-specific networks, comprising modulators, effectors and cognate binding-partners of a specific oncoprotein. SigMaps are reconstructed de novo by integrating diverse evidence sources-including protein structure, gene expression and mutational profiles-via the OncoSig machine learning framework. We first generated a KRAS-specific SigMap for lung adenocarcinoma, which recapitulated published KRAS biology, identified novel synthetic lethal proteins that were experimentally validated in three-dimensional spheroid models and established uncharacterized crosstalk with RAB/RHO. To show that OncoSig is generalizable, we first inferred SigMaps for the ten most mutated human oncoproteins and then for the full repertoire of 715 proteins in the COSMIC Cancer Gene Census. Taken together, these SigMaps show that the cell's regulatory and signaling architecture is highly tissue specific.

Analysis framework and experimental design for evaluating synergy-driving gene expression

The mechanisms by which genetic risk variants interact with each other, as well as environmental factors, to contribute to complex genetic disorders remain unclear. We describe in detail our recently published approach to resolve distinct additive and synergistic transcriptomic effects after combinatorial manipulation of genetic variants and/or chemical perturbagens. Although first developed for CRISPR-based perturbation studies of isogenic human induced pluripotent stem cell-derived neurons, our methodology can be broadly applied to any RNA sequencing dataset, provided that raw read counts are available. Whereas other differential expression analyses reveal the effect of individual perturbations, here we specifically query interactions between two or more perturbagens, resolving the extent of non-additive (synergistic) interactions between perturbations. We discuss the careful experimental design required to resolve synergistic effects and considerations of statistical power and how to quantify observed synergy between experiments. Additionally, we speculate on potential future applications and explore the obvious limitations of this approach. Overall, by interrogating the effect of independent factors, alone and in combination, our analytic framework and experimental design facilitate the discovery of convergence and synergy downstream of gene and/or treatment perturbations hypothesized to contribute to complex diseases. We think that this protocol can be successfully applied by any scientist with bioinformatic skills and basic proficiency in the R programming language. Our computational pipeline ( https://github.com/nadschro/synergy-analysis ) is straightforward, does not require supercomputing support and can be conducted in a single day upon completion of RNA sequencing experiments.

High-throughput toxicogenomic screening of chemicals in the environment using metabolically competent hepatic cell cultures

The ToxCast in vitro screening program has provided concentration-response bioactivity data across more than a thousand assay endpoints for thousands of chemicals found in our environment and commerce. However, most ToxCast screening assays have evaluated individual biological targets in cancer cell lines lacking integrated physiological functionality (such as receptor signaling, metabolism). We evaluated differentiated HepaRGTM cells, a human liver-derived cell model understood to effectively model physiologically relevant hepatic signaling. Expression of 93 gene transcripts was measured by quantitative polymerase chain reaction using Fluidigm 96.96 dynamic arrays in response to 1060 chemicals tested in eight-point concentration-response. A Bayesian framework quantitatively modeled chemical-induced changes in gene expression via six transcription factors including: aryl hydrocarbon receptor, constitutive androstane receptor, pregnane X receptor, farnesoid X receptor, androgen receptor, and peroxisome proliferator-activated receptor alpha. For these chemicals the network model translates transcriptomic data into Bayesian inferences about molecular targets known to activate toxicological adverse outcome pathways. These data also provide new insights into the molecular signaling network of HepaRGTM cell cultures.

Potential Drug Prediction of Glioblastoma Based on Drug Perturbation-Induced Gene Expression Signatures

OBJECTIVES

Glioblastoma (GBM) is a malignant brain tumor which is the most common and aggressive type of central nervous system cancer, with high morbidity and mortality. Despite lots of systematic studies on the molecular mechanism of glioblastoma, the pathogenesis is still unclear, and effective therapies are relatively rare with surgical resection as the frequently therapeutic intervention. Identification of fundamental molecules and gene networks associated with initiation is critical in glioblastoma drug discovery. In this study, an approach for the prediction of potential drug was developed based on perturbation-induced gene expression signatures.

METHODS

We first collected RNA-seq data of 12 pairs of glioblastoma samples and adjacent normal samples from the Gene Expression Omnibus (GEO) database. Differentially expressed genes (DEGs) were identified by DESeq2, and coexpression networks were analyzed with weighted gene correlation network analysis (WGCNA). Furthermore, key driver genes were detected based on the differentially expressed genes and potential chemotherapeutic drugs and targeted drugs were found by correlating the gene expression profiles with drug perturbation database. Finally, RNA-seq data of glioblastoma from The Cancer Genome Atlas (TCGA) dataset was collected as an independent validation dataset to verify our findings.

RESULTS

We identified 1771 significantly DEGs with 446 upregulated genes and 1325 downregulated genes. A total of 24 key drivers were found in the upregulated gene set, and 81 key drivers were found in the downregulated gene set. We screened the Crowd Extracted Expression of Differential Signatures (CREEDS) database to identify drug perturbations that could reverse the key factors of glioblastoma, and a total of 354 drugs were obtained with p value < 10-10. Finally, 7 drugs that could turn down the expression of upregulated factors and 3 drugs that could reverse the expression of downregulated key factors were selected as potential glioblastoma drugs. In addition, similar results were obtained through the analysis of TCGA as independent dataset.

CONCLUSIONS

In this study, we provided a framework of workflow for potential therapeutic drug discovery and predicted 10 potential drugs for glioblastoma therapy.

Lessons from the COVID-19 pandemic for advancing computational drug repurposing strategies

Responding quickly to unknown pathogens is crucial to stop uncontrolled spread of diseases that lead to epidemics, such as the novel coronavirus, and to keep protective measures at a level that causes as little social and economic harm as possible. This can be achieved through computational approaches that significantly speed up drug discovery. A powerful approach is to restrict the search to existing drugs through drug repurposing, which can vastly accelerate the usually long approval process. In this Review, we examine a representative set of currently used computational approaches to identify repurposable drugs for COVID-19, as well as their underlying data resources. Furthermore, we compare drug candidates predicted by computational methods to drugs being assessed by clinical trials. Finally, we discuss lessons learned from the reviewed research efforts, including how to successfully connect computational approaches with experimental studies, and propose a unified drug repurposing strategy for better preparedness in the case of future outbreaks.

Target-Centered Drug Repurposing Predictions of Human Angiotensin-Converting Enzyme 2 (ACE2) and Transmembrane Protease Serine Subtype 2 (TMPRSS2) Interacting Approved Drugs for Coronavirus Disease 2019 (COVID-19) Treatment through a Drug-Target Interaction Deep Learning Model

Previously, our group predicted commercially available Food and Drug Administration (FDA) approved drugs that can inhibit each step of the replication of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) using a deep learning-based drug-target interaction model called Molecule Transformer-Drug Target Interaction (MT-DTI). Unfortunately, additional clinically significant treatment options since the approval of remdesivir are scarce. To overcome the current coronavirus disease 2019 (COVID-19) more efficiently, a treatment strategy that controls not only SARS-CoV-2 replication but also the host entry step should be considered. In this study, we used MT-DTI to predict FDA approved drugs that may have strong affinities for the angiotensin-converting enzyme 2 (ACE2) receptor and the transmembrane protease serine 2 (TMPRSS2) which are essential for viral entry to the host cell. Of the 460 drugs with Kd of less than 100 nM for the ACE2 receptor, 17 drugs overlapped with drugs that inhibit the interaction of ACE2 and SARS-CoV-2 spike reported in the NCATS OpenData portal. Among them, enalaprilat, an ACE inhibitor, showed a Kd value of 1.5 nM against the ACE2. Furthermore, three of the top 30 drugs with strong affinity prediction for the TMPRSS2 are anti-hepatitis C virus (HCV) drugs, including ombitasvir, daclatasvir, and paritaprevir. Notably, of the top 30 drugs, AT1R blocker eprosartan and neuropsychiatric drug lisuride showed similar gene expression profiles to potential TMPRSS2 inhibitors. Collectively, we suggest that drugs predicted to have strong inhibitory potencies to ACE2 and TMPRSS2 through the DTI model should be considered as potential drug repurposing candidates for COVID-19.

Unsupervised and supervised learning with neural network for human transcriptome analysis and cancer diagnosis

Deep learning analysis of images and text unfolds new horizons in medicine. However, analysis of transcriptomic data, the cause of biological and pathological changes, is hampered by structural complexity distinctive from images and text. Here we conduct unsupervised training on more than 20,000 human normal and tumor transcriptomic data and show that the resulting Deep-Autoencoder, DeepT2Vec, has successfully extracted informative features and embedded transcriptomes into 30-dimensional Transcriptomic Feature Vectors (TFVs). We demonstrate that the TFVs could recapitulate expression patterns and be used to track tissue origins. Trained on these extracted features only, a supervised classifier, DeepC, can effectively distinguish tumors from normal samples with an accuracy of 90% for Pan-Cancer and reach an average 94% for specific cancers. Training on a connected network, the accuracy is further increased to 96% for Pan-Cancer. Together, our study shows that deep learning with autoencoder is suitable for transcriptomic analysis, and DeepT2Vec could be successfully applied to distinguish cancers, normal tissues, and other potential traits with limited samples.