scholarly journals Tissue-specific genetic features inform prediction of drug side effects in clinical trials

2020 ◽  
Vol 6 (37) ◽  
pp. eabb6242 ◽  
Author(s):  
Áine Duffy ◽  
Marie Verbanck ◽  
Amanda Dobbyn ◽  
Hong-Hee Won ◽  
Joshua L. Rein ◽  
...  

Adverse side effects often account for the failure of drug clinical trials. We evaluated whether a phenome-wide association study (PheWAS) of 1167 phenotypes in >360,000 U.K. Biobank individuals, in combination with gene expression and expression quantitative trait loci (eQTL) in 48 tissues, can inform prediction of drug side effects in clinical trials. We determined that drug target genes with five genetic features—tissue specificity of gene expression, Mendelian associations, phenotype- and tissue-level effects of genome-wide association (GWA) loci driven by eQTL, and genetic constraint—confer a 2.6-fold greater risk of side effects, compared to genes without such features. The presence of eQTL in multiple tissues resulted in more unique phenotypes driven by GWA loci, suggesting that drugs delivered to multiple tissues can induce several side effects. We demonstrate the utility of PheWAS and eQTL data from multiple tissues for informing drug side effect prediction and highlight the need for tissue-specific drug delivery.

2021 ◽  
Author(s):  
Ghislain Rocheleau ◽  
Iain S Forrest ◽  
Áine Duffy ◽  
Shantanu Bafna ◽  
Amanda Dobbyn ◽  
...  

Background: Phenome-wide association studies conducted in electronic health record (EHR)-linked biobanks have uncovered a large number of genomic loci associated with traits and diseases. However, interpretation of the complex relationships of associated genes and phenotypes is challenging. Results: We constructed a tissue-level phenome-wide network map of colocalized genes and phenotypes. First, we generated colocalized expression quantitative trait loci from 48 tissues of the Genotype-Tissue Expression project and from publicly available genome-wide association study summary statistics from the UK Biobank. We identified 9,151 colocalized genes for 1,411 phenotypes across 48 tissues. Then, we constructed a bipartite network using the colocalized signals to establish links between genes and phenotypes in each tissue. The majority of links are observed in a single tissue whereas only a few are present in all tissues. Finally, we applied the biLouvain clustering algorithm in each tissue-specific bipartite network to identify co-clusters of non-overlapping genes and phenotypes. The majority of co-clusters contains a small number of genes and phenotypes, and 88.6% of co-clusters are found in only one tissue. To demonstrate functionality of the phenome-wide map, we tested if these co-clusters were enriched with known biological and functional gene classes and observed several significant enrichments. Furthermore, we observed that tissue-specific co-clusters are enriched with reported drug side effects for the corresponding drug target genes in clinical trial data. Conclusions: The phenome-wide map provides links between genes, phenotypes and tissues across a wide spectrum of biological classes and can yield biological and clinical discoveries. The phenome-wide map is publicly available at https://rstudio-connect.hpc.mssm.edu/biPheMap/.


2021 ◽  
Author(s):  
Hao Lu ◽  
Luyu Ma ◽  
Lei Li ◽  
Cheng Quan ◽  
Yiming Lu ◽  
...  

Noncoding genomic variants constitute the majority of trait-associated genome variations; however, identification of functional noncoding variants is still a challenge in human genetics, and a method systematically assessing the impact of regulatory variants on gene expression and linking them to potential target genes is still lacking. Here we introduce a deep neural network (DNN)-based computational framework, RegVar, that can accurately predict the tissue-specific impact of noncoding regulatory variants on target genes. We show that, by robustly learning the genomic characteristics of massive variant-gene expression associations in a variety of human tissues, RegVar vastly surpasses all current noncoding variants prioritization methods in predicting regulatory variants under different circumstances. The unique features of RegVar make it an excellent framework for assessing the regulatory impact of any variant on its putative target genes in a variety of tissues. RegVar is available as a webserver at http://regvar.cbportal.org/.


2020 ◽  
Author(s):  
Maud Fagny ◽  
Marieke Lydia Kuijjer ◽  
Maike Stam ◽  
Johann Joets ◽  
Olivier Turc ◽  
...  

AbstractEnhancers are important regulators of gene expression during numerous crucial processes including tissue differentiation across development. In plants, their recent molecular characterization revealed their capacity to activate the expression of several target genes through the binding of transcription factors. Nevertheless, identifying these target genes at a genome-wide level remains a challenge, in particular in species with large genomes, where enhancers and target genes can be hundreds of kilobases away. Therefore, the contribution of enhancers to regulatory network is still poorly understood in plants. In this study, we investigate the enhancer-driven regulatory network of two maize tissues at different stages: leaves at seedling stage and husks (bracts) at flowering. Using a systems biology approach, we integrate genomic, epigenomic and transcriptomic data to model the regulatory relationship between transcription factors and their potential target genes. We identify regulatory modules specific to husk and V2-IST, and show that they are involved in distinct functions related to the biology of each tissue. We evidence enhancers exhibiting binding sites for two distinct transcription factor families (DOF and AP2/ERF) that drive the tissue-specificity of gene expression in seedling immature leaf and husk. Analysis of the corresponding enhancer sequences reveals that two different transposable element families (TIR transposon Mutator and MITE Pif/Harbinger) have shaped the regulatory network in each tissue, and that MITEs have provided new transcription factor binding sites that are involved in husk tissue-specificity.SignificanceEnhancers play a major role in regulating tissue-specific gene expression in higher eukaryotes, including angiosperms. While molecular characterization of enhancers has improved over the past years, identifying their target genes at the genome-wide scale remains challenging. Here, we integrate genomic, epigenomic and transcriptomic data to decipher the tissue-specific gene regulatory network controlled by enhancers at two different stages of maize leaf development. Using a systems biology approach, we identify transcription factor families regulating gene tissue-specific expression in husk and seedling leaves, and characterize the enhancers likely to be involved. We show that a large part of maize enhancers is derived from transposable elements, which can provide novel transcription factor binding sites crucial to the regulation of tissue-specific biological functions.


Blood ◽  
2012 ◽  
Vol 120 (21) ◽  
pp. 528-528
Author(s):  
Huimin Geng ◽  
Rahul Nahar ◽  
Parham Ramezani-Rad ◽  
Zhengshan Chen ◽  
Jeffrey W. Tyner ◽  
...  

Abstract Abstract 528 Background: Tyrosine kinase-driven B-cell lineage acute lymphoblastic leukemia such as Ph+ ALL, is associated with particularly poor clinical outcome and defined by deletions of the IKZF1 gene encoding the Ikaros transcription factor in >80% of the cases. While the role of Ikaros in lymphoid differentiation has been extensively studied, it remains elusive how Ikaros functions as tumor suppressor in Ph+ALL. Here we demonstrate that Ikaros orchestrates a set of gene expression changes that affects leukemia cell proliferation and survival at multiple levels. Results: A comprehensive gene expression analysis revealed that reconstitution of Ikaros function in patient-derived Ph+ ALL cells mimics gene expression changes induced by Pax5 (a B cell specific transcription factor), μ-chain and BLNK (pre-B cell receptor components), and treatment with Imatinib (a BCR-ABL1 kinase inhibitor), while it reverses Myc- and Stat5a/b-dependent survival signals. In combining seven experimental perturbations (reconstitution of Ikzf1, Pax5, μ-chain, Blnk, treatment with Imatinib or deletion of Stat5a/bfl/fl and Mycfl/fl), we identified a commonly regulated gene set, including known tumor suppressors SPIB, BTG1, and BTG2 and highly enriched for Myc-related gene signatures. These findings suggest that reconstitution of PAX5 and IKZF1 transcription factors converges with pre-B cell receptor-mediated transcriptional changes. We next connected our analysis at the transcriptional level with Ikaros-dependent changes at the level of tyrosine phosphorylation events: The BCR-ABL1 kinase in Ph+ ALL drives proliferation and survival through phosphorylation of Stat5, AKT, ERK, SRC and multiple other signaling intermediates. A global mass spectrometry analysis of phosphotyrosine proteins in patient-derived Ph+ ALL cells showed that reconstitution of Ikaros leads to global loss of phospho-tyrosine, in a comparable manner as treatment with TKI. The vast majority of Ikaros-mediated changes resulted in loss of tyrosine phosphorylation such as the SRC tyrosine kinase LCK (Y504 and Y192), the proximal BCR-ABL1 adapter molecule NEDD9 (Y166, Y240, Y344 and Y578) and ABL1 itself (Y257). For a small number of molecules, Ikaros-reconstitution resulted in increased tyrosine phosphorylation including the phosphatases Shp2 (Ptpn11Y584) and Ship2 (Inppl1Y1161). A focus phospho-tyrosine analysis of individual signaling molecules analysis revealed that Stat5, SRC, AKT and ERK are all affected by Ikaros-induced de-phosphorylation. These results suggest that that Ikaros induces dephosphorylation of multiple signaling intermediates in BCR-ABL1 signaling. To identify Ikaros-mediated gene expression changes that were directly induced by Ikaros promoter binding, we matched Ikaros-dependent gene expression changes with Ikaros ChIP-seq data. This analysis confirmed enriched binding of Ikaros on Btg1, Myc and Smyd2 promoters. Among Ikaros target genes, we identified a set of genes that predicts clinical outcome for adult and pediatric patients with ALL. According to this predictor, IkarosUP (FBXO33, HBP1, HIVEP1, DTX1, KIF13B, RHOQ) and IkarosDOWN (PFKP, ASNS, ETV6, PSTPIP2, SMYD2, CDKL5), patients with B-ALL were segregated into two groups with significantly different outcomes, with a 5-year RFS of 76% vs 29% (COG P9906 trial, n=110; p=0.006) and a 5-year OS of 39% vs 0% (ECOG E2993 trial, n=25; p=0.007). Further biochemistry and mouse experiments showed that both the ability of Ikaros to dephosphorylate Stat5 and to function as tumor suppressor requires cooperation with the pre-B cell receptor linker molecule BLNK but not the proximal effector kinase Syk. Conclusions: Importantly, Ikaros binds to and transcriptionally activates CDKN2A (Arf) and TP53 loci and thereby opposes the function of BCL6, which acts as transcriptional repressor at these loci. Comparing the gene expression microarray and ChIP-seq or ChIP-on-Chip data of BCL6 and Ikaros, we identified a set of target genes that is regulated by both BCL6 and Ikaros and found that for the vast majority of Ikaros/BCL6 target genes, Ikaros-reconstitution recapitulated a BCL6−/− situation, i.e. genes that are induced by Ikaros are repressed by BCL6. These findings are of potential clinical relevance since pharmacological inhibition of BCL6 restores sensitivity of IKZF1-mutant Ph+ ALL cells to conventional tyrosine kinase inhibitors. Disclosures: Druker: ARIAD: OHSU receives clinical trial funding. Dr. Druker is currently principal investigator or co-investigator on Novartis, Bristol-Myers Squibb, and ARIAD clinical trials. His institution has contracts with these companies to pay for patient costs, nurse and da Other; Bristol-Myers Squibb: OHSU receives clinical trial funding, OHSU receives clinical trial funding Other; Novartis: OHSU receives clinical trial funding, OHSU receives clinical trial funding Other; MolecularMD: Consultancy, Equity Ownership, Membership on an entity's Board of Directors or advisory committees, Scientific Founder; OHSU and Dr. Druker have a financial interest in MolecularMD. OHSU has licensed technology used in some of these clinical trials to MolecularMD. This potential individual and institutional conflict of interest has been reviewed and man, Scientific Founder; OHSU and Dr. Druker have a financial interest in MolecularMD. OHSU has licensed technology used in some of these clinical trials to MolecularMD. This potential individual and institutional conflict of interest has been reviewed and man Other.


2019 ◽  
Author(s):  
Diego Galeano ◽  
Alberto Paccanaro

AbstractPair-input associations for drug-side effects are obtained through expensive placebo-controlled experiments in human clinical trials. An important challenge in computational pharmacology is to predict missing associations given a few entries in the drug-side effect matrix, as these predictions can be used to direct further clinical trials. Here we introduce the Geometric Sparse Matrix Completion (GSMC) model for predicting drug side effects. Our high-rank matrix completion model learns non-negative sparse matrices of coefficients for drugs and side effects by imposing smoothness priors that exploit a set of pharmacological side information graphs, including information about drug chemical structures, drug interactions, molecular targets, and disease indications. Our learning algorithm is based on the diagonally rescaled gradient descend principle of non-negative matrix factorization. We prove that it converges to a globally optimal solution with a first-order rate of convergence. Experiments on large-scale side effect data from human clinical trials show that our method achieves better prediction performance than six state-of-the-art methods for side effect prediction while offering biological interpretability and favouring explainable predictions.


2016 ◽  
Author(s):  
Farhad Hormozdiari ◽  
Martijn van de Bunt ◽  
Ayellet V. Segrè ◽  
Xiao Li ◽  
Jong Wha J Joo ◽  
...  

AbstractThe vast majority of genome-wide association studies (GWAS) risk loci fall in non-coding regions of the genome. One possible hypothesis is that these GWAS risk loci alter the individual’s disease risk through their effect on gene expression in different tissues. In order to understand the mechanisms driving a GWAS risk locus, it is helpful to determine which gene is affected in specific tissue types. For example, the relevant gene and tissue may play a role in the disease mechanism if the same variant responsible for a GWAS locus also affects gene expression. Identifying whether or not the same variant is causal in both GWAS and eQTL studies is challenging due to the uncertainty induced by linkage disequilibrium (LD) and the fact that some loci harbor multiple causal variants. However, current methods that address this problem assume that each locus contains a single causal variant. In this paper, we present a new method, eCAVIAR, that is capable of accounting for LD while computing the quantity we refer to as the colocalization posterior probability (CLPP). The CLPP is the probability that the same variant is responsible for both the GWAS and eQTL signal. eCAVIAR has several key advantages. First, our method can account for more than one causal variant in any loci. Second, it can leverage summary statistics without accessing the individual genotype data. We use both simulated and real datasets to demonstrate the utility of our method. Utilizing publicly available eQTL data on 45 different tissues, we demonstrate that computing CLPP can prioritize likely relevant tissues and target genes for a set of Glucose and Insulin-related traits loci. eCAVIAR is available at http://genetics.cs.ucla.edu/caviar/


2021 ◽  
Vol 12 (1) ◽  
Author(s):  
Charleen Hunt ◽  
Suzanne A. Hartford ◽  
Derek White ◽  
Evangelos Pefanis ◽  
Timothy Hanna ◽  
...  

AbstractCRISPR-based transcriptional activation is a powerful tool for functional gene interrogation; however, delivery difficulties have limited its applications in vivo. Here, we created a mouse model expressing all components of the CRISPR-Cas9 guide RNA-directed Synergistic Activation Mediator (SAM) from a single transcript that is capable of activating target genes in a tissue-specific manner. We optimized Lipid Nanoparticles and Adeno-Associated Virus guide RNA delivery approaches to achieve expression modulation of one or more genes in vivo. We utilized the SAM mouse model to generate a hypercholesteremia disease state that we could bidirectionally modulate with various guide RNAs. Additionally, we applied SAM to optimize gene expression in a humanized Transthyretin mouse model to recapitulate human expression levels. These results demonstrate that the SAM gene activation platform can facilitate in vivo research and drug discovery.


Blood ◽  
2005 ◽  
Vol 106 (11) ◽  
pp. 740-740
Author(s):  
Teresa Palomero ◽  
Duncan T. Odom ◽  
Adam Margolin ◽  
Jon Aster ◽  
Andrea Califano ◽  
...  

Abstract The NOTCH1 signaling pathway plays a critical role in the regulation of hematopoietic stem cell homeostasis and is required for lymphocytes to adopt a T-cell fate. Importantly, aberrant activation of NOTCH signaling due to activating mutations in the NOTCH1 gene is involved in the pathogenesis of over 50% human T-cell acute lymphoblastic leukemias (T-ALL) and the therapeutic efficacy of inhibition of NOTCH signaling via gamma-secretase inhibitors (GSI) is currently being tested in clinical trials. However, little is known about the transcriptional programs activated downstream of NOTCH1 activation that contribute to the transformation of T-cell progenitors. Here we used chromatin immunoprecipitation and promoter microarrays (ChIP-on-chip) to identify direct transcriptional targets of NOTCH1; and microarray gene expression analysis to decipher the oncogenic transcriptional network activated by the NOTCH1 oncoprotein in T-ALL cells. Using the Hu19K arrays that contain over 13,000 human promoter regions, we have identified 134 candidate direct targets (p<0.0001) of NOTCH1 in the HPB-ALL T-ALL cell line by ChIP-on-chip. NOTCH1 targets identified by ChIP -on-chip included known direct targets of NOTCH1 such as preTCRA, as well as genes involved in T-cell signaling (CD3D, ASE-1/CD3E associated protein), proliferation (CDK5, CDC25A, RBL1, BUB3 and ING3) and survival (Survivin). Inhibition of NOTCH1 signaling with compound E, a strong GSI, induced a gene expression signature characterized by the downregulation of known NOTCH1 direct target genes such as DELTEX1, HES1 and HEY1. Analysis of different cell lines representative of different stages of maturation arrest and different oncogenic groups of T-ALL identified a core signature of transcriptional responses to NOTCH signaling inhibition, which included the downregulation of the MYC oncogene as well as numerous genes involved in nucleotide metabolism and protein synthesis. In addition, cell cycle inhibitors p27/KIP1 and p18/INK4D were upregulated. Importantly, 13 of the top direct target genes of NOTCH1 identified by ChIP-on-chip were consistently downregulated (p<0.005) upon NOTCH signaling inhibition in multiple cell lines. However, more restricted responses to NOTCH1 signaling inhibition, which included important developmental regulators of T-cell development such as the pre-TCRA, were present in specific groups of samples. These results demonstrate that NOTCH1 activation induces a complex transcriptional response in T-ALL cells, which is in part dependent on the stage of T-cell development and/or the interaction with other T-ALL transcription factor oncogenes. Common effector pathways downstream of NOTCH signaling may represent novel therapeutic targets for the treatment of T-ALL, while cell type specific responses may influence the cellular effects and the clinical efficacy of NOTCH signaling inhibitors currently under evaluation in clinical trials. Overall, our results indicate that NOTCH1 acts as a master transcriptional regulator at the top of a complex regulatory network that contributes to leukemogenesis by regulating multiple critical pathways involved in the regulation of T-cell differentiation, proliferation and survival.


2013 ◽  
Vol 31 (31_suppl) ◽  
pp. 17-17
Author(s):  
Praveen Ramakrishnan Geethakumari ◽  
Joann R. Ackler ◽  
Mahasweta Gooptu ◽  
Leonard E Braitman ◽  
William J. Tester

17 Background: Clinical trials are fundamental to innovative oncology. Participation rates in trials have declined nationally to < 5%. Barriers to participation exist at patient, physician, and protocol levels. This study seeks to identify barriers to enrollment in clinical trials at a community cancer center serving a diverse patient population. Methods: We conducted a descriptive cross-sectional study, including 160 eligible patients offered enrollment in 27 clinical trials from July 2010-December 2012. A standardized questionnaire was delivered by mail or in person. Patients who enrolled (acceptors) and decliners were compared using Fisher’s exact test for nominal variables and t test for normally distributed continuous variables. Results: Fifty-seven patients (36%) (males: 10, females: 47) returned the questionnaire. Thirty-three (58%) were enrolled in a clinical trial. Mean age of acceptors was 57 compared to 64 for decliners (p=0.007). Stage IV disease patients were more likely to enroll (Spearman rho= 0.33, p=0.01). Among patients with family support, 66% accepted participation compared to 40% of those without [p=0.05]. Twenty-eight of 33 (85%) who felt trust in their doctor affected their decision enrolled. Of those “comfortable with randomization,” 86% enrolled compared to 29% who were not [p<0.001]. 74% patients would participate in a trial if maximal information could be gathered before making a final decision. Acceptors stated altruism, contribution to research, trust in doctor and hope for cure while decliners mentioned uncertainty in research, drug side effects, mistrust in pharmaceutical industry and depression as most important reasons for their decision. Conclusions: Our results confirm barriers among diverse patients treated at a community cancer center. Study limitations include small sample size and predominance of female gender. Factors influencing enrollment identified include age, family support, patient’s insight into conduct of randomized trials, perceived drug side effects and the doctor-patient relationship. Success lies in bridging knowledge and communication gaps, careful protocol design, and establishing trusting relationships.


2014 ◽  
Vol 96 ◽  
Author(s):  
KAN HE ◽  
JIAOFANG SHAO ◽  
ZHONGYING ZHAO ◽  
DAHAI LIU

SummaryThe fundamental step of learning transcriptional regulation mechanism is to identify the target genes regulated by transcription factors (TFs). Despite numerous target genes identified by chromatin immunopre-cipitation followed by high-throughput sequencing technology (ChIP-seq) assays, it is not possible to infer function from binding alone in vivo. This is equally true in one of the best model systems, the nematode Caenorhabditis elegans (C. elegans), where regulation often occurs through diverse TF binding features of transcriptional networks identified in modENCODE. Here, we integrated ten ChIP-seq datasets with genome-wide expression data derived from tiling arrays, involved in six TFs (HLH-1, ELT-3, PQM-1, SKN-1, CEH-14 and LIN-11) with tissue-specific and four TFs (CEH-30, LIN-13, LIN-15B and MEP-1) with broad expression patterns. In common, TF bindings within 3 kb upstream of or within its target gene for these ten studies showed significantly elevated level of expression as opposed to that of non-target controls, indicated that these sites may be more likely to be functional through up-regulating its target genes. Intriguingly, expression of the target genes out of 5 kb upstream of their transcription start site also showed high levels, which was consistent with the results of following network component analysis. Our study has identified similar transcriptional regulation mechanisms of tissue-specific or broad expression TFs in C. elegans using ChIP-seq and gene expression data. It may also provide a novel insight into the mechanism of transcriptional regulation not only for simple organisms but also for more complex species.


Sign in / Sign up

Export Citation Format

Share Document