A Web Application for Competing Endogenous RNA Exploration
Gene expression silencing at mRNA level by microRNAs is a well-established form of post-transcriptional regulation.1,2 Such silencing is achieved through microRNA binding to microRNA response elements (MREs) residing mainly in the 3′ untranslated regions of the target mRNA. Over 1000 human microRNAs have been identified,3 and the prevalence of microRNA regulation in a broad range of biological processes and disease often attributes to the fact that a single microRNA can repress hundreds of different mRNAs.4 Interestingly, a single target mRNA often possesses MREs of distinct microRNAs in its 3′ untranslated regions.5 Questions have been raised regarding the need for this redundancy in regulation, and these multiple MREs were once thought to serve as regulatory buffers of different microRNAs. In a recent seminal work, a novel theory, termed the competing endogenous RNA (ceRNA), was proposed to provide a plausible explanation for this interesting phenomenon from a new perspective of gene regulation.6 According to the ceRNA theory, MREs function as letters of this new regulatory system, and ceRNAs, or sets of RNAs including mRNA, pseudogenes, and long noncoding RNAs, can communicate or regulate each other, through competition for common MREs. As such, ceRNA regulatory networks provide a unifying system for regulations among transcriptome-wide RNAs, greatly expanding the functions of RNAs.6 Alteration of this competition between ceRNAs could modify normal state gene expression and in return alter the status of biological pathways to promote an oncogenic program, for example. To that end, a PTEN ceRNA network was uncovered and shown to potentially regulate oncogenesis.6
The fact that this new level of RNA regulations could be prevalent in cells has prompted research to identify ceRNAs of genes related to disease. However, the complexity of ceRNA regulations and an incomplete knowledge of microRNA binding have hampered the prediction of ceRNAs, which often requires the use of computational tools and databases that are not readily available to the users. Thus far, 2 algorithms for human ceRNA predictions have been proposed. MuTaMe6 aims to predict ceRNAs of a gene of interest (GOI). It starts by selecting a set of, ideally experimentally validated, microRNAs that target the given GOI in its 3′ untranslated regions. Predicted ceRNAs by sequence pairing are the mRNAs that are also targeted by these microRNAs, and the prediction is made based on scores generated from binding affinity statistics. Although MuTaMe succeeded in predicting several ceRNAs of PTEN, it is not accessible for predicting other GOIs because experimentally validated microRNAs targeting a new GOI are mostly unavailable, and binding affinity statistics used in MuTaMe are insufficient for accurate predictions. Furthermore, MuTaMe has not been implemented as a software tool yet and cannot be accessed by the general public. The second algorithm, Hermes,7 infers ceRNAs from expression profiles of genes and microRNAs by using conditional mutual information. Although Hermes combines ceRNA/microRNA/target triplets via tissue-specific gene expression, however, it does not provide an implementation that combines sequence binding statistics with gene expression. There is a shortage of user-friendly tools that can be easily used for anyone interested in ceRNA research.
To address the need for user-friendly tools, we developed TraceRNA, a web-based application for transcriptome-wide ceRNA discovery. TraceRNA is flexible, powerful, and user-friendly. It includes MiRTarBase,8 a database of experimentally validated microRNA target pairs and microRNA binding scores and related data (site position, length, etc) from 3 prediction algorithms (SVMicrO,9 BCMicrO,10 and SiteTest) with different emphasis. TraceRNA provides the user with the flexibility to perform ceRNA predictions using 1 of 3 algorithms to meet different study objectives. Currently, TraceRNA maintains a database that includes genome-wide targets of >700 human microRNAs predicted by 3 algorithms. The user can compare among the prediction results from these different algorithms to either complement or reach a consensual prediction.
Two important observations have been integrated into the TraceRNA for context-specific ceRNA discovery. The first is that the microRNA expression is condition-specific. That is, if a microRNA is not expressed in a tissue environment or disease state, one can ignore its target-binding specificity. The second is that GOI and its ceRNAs’ expressions are positively correlated because of the competition for microRNA binding. Therefore, an increased/decreased GOI expression will attract more/less microRNA binding away from its ceRNAs, resulting in increased/decreased ceRNA expression level because of the decreased/increased repression effect of microRNAs. As another unique feature, TraceRNA can construct ceRNA interaction networks to help delineate complex interactions of ceRNAs and gain further insight into this novel ceRNA regulation– modulation mechanism. Finally, TraceRNA is developed to be a user-friendly web application with an accessible interface. It generates predictions including statistics such as P values and false discovery rate both online and in spreadsheets available for download.
The goal of TraceRNA is to predict ceRNAs of a GOI, which are mRNAs that share MREs from a set of microRNAs that also target the GOI. In this article, we named these microRNAs as GOI-targeting microRNAs (GTmiRs). ceRNAs’ competition for GTmiR binding to GOI will alter the expression of GOI and its ceRNAs in a coordinated fashion, and coexpression can be observed, where expressions of GOI and its ceRNAs are expected to be correlated. Predictions of ceRNAs can be done by examining microRNA–mRNA sequence pairing or GOI–ceRNA coexpression. TraceRNA includes 3 main processing sections in its pipeline (Figure 1): (1) sequence-based prediction of ceRNAs, (2) coexpression analysis of GOI and ceRNAs’ expression levels, and (3) generation of ceRNA regulatory network. Additionally, microRNA expression data are also included in TraceRNA for the user to select context-specific GTmiRs (Figure I in the Data Supplement).
Sequence-Based Prediction of ceRNAs
Selection of GTmiRs
Given a GOI provided by the user, the first step in TraceRNA is to identify GTmiRs. TraceRNA provides 2 alternatives for GTmiR identification (Figure 1). First, TraceRNA maintains a local copy of experimentally validated microRNAs: target pairs curated by miRTarBase release 2.5 (downloaded on July 2012). Second, genome-wide SVMicrO9 predictions for >700 microRNAs were precalculated. SVMicrO9 was developed previously to predict microRNA targets. It uses a support vector machine with sequence-based features, including binding secondary structure, energy, binding conservation, number of predicted sites, and site densities. SVMicrO was tested to achieve improved performance compared with several popular algorithms, including TargetScan, miRanda, and Pictar. The predicted microRNAs are displayed to the user in descending order of P values. The user may select a subset or all of the microRNAs from these 2 sources as GTmiRs.
Prediction of ceRNAs
Once GTmiRs are selected, TraceRNA predicts ceRNAs as the mRNAs that are also targeted by these GTmiRs, by using 1 of 3 microRNA target prediction algorithms, SVMicrO,9 BCMicrO,10 or SiteTest, depending on the user’s selection. SVMicrO9 and BCMicrO10 are 2 inhouse-developed algorithms, which were published previously. As discussed above, SVMicrO makes predictions by using a large number of microRNA-binding features. BCMicrO uses a Bayesian approach that integrates prediction scores from 6 popular algorithms: TargetScan,11 miRanda,12 PicTar,13 mirTarget2,14 PITA,15 and DIANA micro-T.16 Both algorithms provide more accurate predictions than existing algorithms. The prediction scores of SVMicrO and BCMicrO were precalculated and stored in a MySQL database. In addition, a new algorithm, SiteTest, inspired by MuTaMe,6 was also developed, and its pseudo code is included in the Data Supplement.
To show the score calculation, let Si be the score of GTmiR i targeting an mRNA by either algorithms and K be the total number of GTmiRs. Then, the score, S, for the mRNA to be a ceRNA predicted by sequence pairing is calculated as(1)
We discuss next the calculation of the predictions significance.
Statistical Significance of Predicted ceRNAs
We first discuss the calculation of statistical significance for the SVMicrO scores. According to Equation 1, S is calculated as the average of the sequence pairing scores of each GTmiR and the mRNA. To calculate the P-value for S, the distribution of S under the null hypothesis, that is, the mRNA is not predicted by sequence pairing as a ceRNA, needs to be obtained. Because S is the average Si, then the distribution of Si under the null hypothesis needs to be evaluated first. Adopting the method developed in BCMicrO, the empirical distribution Si under the null hypothesis was observed as a mixture of 2 distributions, one clustered around smaller scores and the other around larger scores (Figure 2). Given that most genes are not microRNA targets and they should have smaller SVMicrO scores, we hypothesized that the distribution around smaller scores characterizes the scores derived from genes not targeted by any microRNA, which was further assumed to follow the independent identically distributed gamma distribution, or Si ≈ Gamma (–,†) whose parameters α and β were obtained from fitting the empirical scores Si (Figure I in the Data Supplement). Subsequently, because of Equation 1, S is also gamma-distributed under the null hypothesis as:(2)
Therefore, the probability (P value) of a sequence pairing prediction score S can be evaluated analytically by Equation 2. The same method was applied to BCMicrO and SiteTest by fitting the gamma distributions directly to their scores. Once P values of all predicted ceRNAs by sequence pairing are calculated, the corresponding false discovery rates are computed using the Benjamini–Hochberg method.17
Coexpression-Based Prediction of ceRNAs
Test for Coexpression Between GOI and Predicted ceRNAs by Sequence Pairing
TraceRNA can also integrate a tissue- or disease-specific expression data set to predict tissue- or disease-specific ceRNAs of the GOI and potentially further improve the prediction specificity (Figure 1). Currently, expression data sets of glioblastoma multiforme (GBM)18 and breast cancer19 from The Cancer Genome Atlas (TCGA; (http://cancergenome.nih.gov/) are included. The users may contact the webmaster to upload their own expression data sets if needed. Because higher GOI expression competitively attracts more microRNA binding and thus reduces the possibility of the same microRNA binding to ceRNAs, leading to higher ceRNA expression, the coexpression analysis first computes the Pearson correlation coefficients between GOI expression levels and predicted ceRNAs by sequence pairing and then removes the mRNAs with negative correlation coefficients. The P values were calculated by Fisher transformation20 and the resultant predictions have 2 scores: those by sequence pairing and those by coexpression test. We discuss their consolidation in the next section.
To fuse these 2 scores, we used the Borda counting method,21 which essentially sums ranks of scores. The resultant ceRNA list can be downloaded from the Website as a common delimited text file that contains the gene symbols ordered based on the Borda scores from highest to lowest, their sequence pairing scores, coexpression scores, and their rankings.
Generation of Regulatory Network Based on a GOI
TraceRNA also aims to provide a tool that allows biologists to discover new regulatory networks that are potentially modulated by a set of GTmiRs and gain insight into this novel gene regulation–modulation mechanism. To generate a GOI–ceRNA regulatory network, the user can select top predicted ceRNAs by coexpression test for a given GOI and then treat each predicted ceRNA as a new GOI (or cGOI). TraceRNA performs new rounds of predictions for each cGOI iteratively using the same number of predicted microRNAs that target each cGOI as described before. The resulting list (containing GOI, ceRNAs, and scores for all cGOIs) is used to generate a regulation network using Cytoscape plug-in,22 which can be downloaded for further analysis.
Biological Functional Enrichment
To examine the functional association of ceRNAs for a given GOI, we used DAVID23 (http://david.abcc.ncifcrf.gov/), which uses a modified Fisher exact test to evaluate the functional enrichment of 40 annotation categories, including GO terms, protein–protein interactions, disease associations, pathways, homologies, and other gene sets in a given gene list. In this article, the enrichment results for P value <0.01 are reported.
Final Remarks on Methods
Discussion of TraceRNA implementation can be found in Data Supplement. Table 1 summarizes the algorithms and databases used in TraceRNA. MiRTarBase, SVMicrO, and BCMicrO were implemented as databases queried by SQL commands. SiteTest accesses SVMicrO database and calculates binding scores for each ceRNA. All the computations, including statistical significance and Borda fusion, were implemented by R (http://www.r-project.org/).
Results and Case Studies
TraceRNA integrates databases, SQL queries, real-time predictions, and generation of ceRNA interaction network under a unified Web interface, enabling ceRNA predictions and discovery of novel biological regulation. We illustrate its features and capabilities next.
TraceRNA Web Interface
The TraceRNA Web interface (Figure I in the Data Supplement) starts with a query GOI by the user. Currently, TraceRNA only supports official gene symbols from the University of California Santa Cruz annotation. Given a GOI, a set of validated microRNAs that target the GOI derived from miRTarBase will be displayed under the checker box Select Validated miRNAs. Other microRNAs predicted by SVMicrO are listed under Select Predicted miRNAs in an increasing order of binding P-values of targeting GOI. The user can select from these 2 sources a set of microRNAs to form GTmiRs. Here is a rule of thumb: one can require a binding P value<0.01, which is expected to produce <43 microRNAs for 50% of genes (Figure II in the Data Supplement), or microRNAs log2 expression level in GBM >6 (Figure III in the Data Supplement).
After selection, the user can choose from SVMicrO, BCMicrO, or SiteTest and further integrate gene expression data. To evaluate ceRNA-mediated gene–gene interactions, the user can perform ceRNA prediction iteratively by treating top K (20 by default) ceRNAs as GOI (cGOI). The resulting interactions will be displayed within the Web interface and can be also saved in a file to be imported into Cytoscape.
Determination of GTmiRs that target the GOI is an important step that can significantly affect the final prediction. GTmiRs are ideally determined by experiments, but the complete set of experimentally validated GTmiRs is rarely available. ceRNA predictions based on a subset of validated GTmiRs will have low specificity, and predicted GTmiRs are needed to increase the specificity. However, high false positives associated with current microRNA target prediction algorithms could introduce false positives in ceRNA predictions, thus potentially harming rather than improving the ceRNA prediction specificity. Furthermore, a different number of candidate GTmiRs can also affect the prediction performance, where a lower number will likely produce lower prediction specificity, whereas too high a number can harm the prediction sensitivity. This observation was captured in Figure IV in the Data Supplement, in which we varied the number of microRNAs from 2 to 70. It clearly demonstrated the low specificity with a small number of microRNAs and a loss of sensitivity when too many microRNAs were selected. Therefore, care needs to be taken in choosing GTmiRs. To this end, TraceRNA provides flexibility to choose between validated and predicted GTmiRs or a combination of both.
As an example, Table 2 includes the top 20 experimentally validated (from miRTarBase) and predicted GTmiRs for PTEN, ESR1, and BRCA1, respectively. In all 3 cases, the numbers of experimentally validated GTmiRs are <20, and there are only 4 for BRCA1. Apparently, using the validated GTmiRs alone will result in low specificity in ceRNA predictions.
Significance of ceRNA Prediction Score
Figure 2A depicts the empirical distribution of the genome-wide SVMicrO scores for 772 human microRNAs, and a mixture of 2 distributions can be clearly observed, one clustered around smaller scores with a much larger mass and the other around larger scores. As discussed in the Methods section, the peak around smaller scores was considered to represent the null distribution and was fitted with a gamma distribution, whose parameters are α=0.7234 and β=0.3594 with 95% confidence interval (0.7229, 0.7239) and (0.3591, 0.3598) for α and β, respectively. Figure 2B shows the histogram and the fitted gamma distribution (with a constant shift). Table I in the Data Supplement lists the fitted parameters of the gamma distributions for SVMicrO, BCMicrO, and SiteTest.
Case Study 1
We applied TraceRNA to predict the ceRNAs of PTEN. PTEN is a gene related to the development of many cancers, where it often functions as a key tumor suppressor, whose abundance determines the critical outcomes in tumorigenesis.6 PTEN is also known to regulate cell cycle, particularly in preventing cells from growing and dividing too rapidly. PTEN ceRNAs have also been predicted and reported.6,7
To predict ceRNAs of PTEN with TraceRNA, we selected only the predicted microRNAs as GTmiRs (Table 2) and further chose SVMicrO to predict the ceRNAs by sequence pairing. TraceRNA returned a total of 761 predicted ceRNAs by sequence pairing for P value <0.05, and the 20 best predictions (in descending order of prediction score) together with P values are provided in Table 3. One important feature of TraceRNA is the possibility to increase the predictions of specificity and predict context-specific ceRNAs by integrating an expression data set of a disease condition. In this case, 400 GBM expression samples from TCGA project were included, based on which coexpression correlations against PTEN were evaluated and a total of 466 genes were obtained for Pearson correlation greater than zero as the GBM-specific ceRNAs of PTEN. Table 3 shows the results for the top 20 predicted ceRNAs based on Borda score. To examine their functions, pathway enrichment was performed using DAVID23 on these 466 ceRNAs, and the 3 enriched pathways are shown in Table 4 (under TraceRNA plus GBM). Well-known cancer-related pathways, including mitogen-activated protein kinase signaling and Wnt signaling, were significantly enriched, indicating an important involvement of PTEN ceRNAs in cancer. Another enriched pathway, transforming growth factor-β signaling, is known to use intracellular SMADs to mediate growth suppression and PTEN downregulation simultaneously to induce growth proliferation. Here, the prediction result provided a third possible regulatory mechanism of transforming growth factor-β regulation by PTEN via its ceRNAs.
It would be also interesting to examine if these GBM-specific ceRNAs by coexpression test also have higher specificity than those by sequence pairing alone (Table 3). However, direct comparison was infeasible because of a lack of true PTEN ceRNAs. Alternatively, pathway enrichment of the prediction result was conducted to make an indirect comparison. Intuitively, true ceRNAs should be functionally more significant than false-positive predictions, and therefore, the predictions with higher specificity should be accompanied by a larger number of more enriched pathways. Pathway enrichment of sequence-based predictions is shown in Table 4 (under TraceRNA), and it is apparent that the GBM-specific ceRNAs by coexpression test are of higher functional enrichment (9 enriched functions versus 4 weakly enriched functions in sequence pairing predictions alone), thus a higher prediction specificity. Expression (GBM data) scatterplots of PTEN versus 3 predicted ceRNAs (QKI, NOVA1, and BCL11A) by sequence predictions (Table 3) are shown in Figure 3A to 3C. The correlations are clearly very low, suggesting that they are not GBM-specific ceRNAs. As expected, they were not among the predicted GBM-specific ceRNAs (Table 3). Expression scatterplots of PTEN versus the 3 predicted GBM-specific ceRNAs (GSPT1, PPP6C, and USP15; Table 3) are shown in Figure 3D to 3F. Their correlations are much higher. Notice that USP15 was also ranked sixth in sequence-based predictions. As expected, its ranking improved after integrating gene expression data.
As a comparison, ceRNAs predicted by Competitive Endogenous mRNA DataBase (ceRDB)24 were also retrieved and top 20 are listed in Table 3. We observed only 1 overlap between ceRDB and TraceRNA predictions in the top 20 predictions. To examine the functional significance of ceRDB predictions, pathway enrichment was conducted (Table 4, column ceRDB). Nine pathways in TraceRNA plus GBM are enriched compared with 5 pathways in ceRDB. For example, enrichment P values of Long-term Potentiation that includes genes such as MAPK1, NRAS, RPS6KA3, KRAS, and CREBBP are 8.8×10–8, 0.00002, and 0.0084 for TraceRNA plus GBM, TraceRNA sequence pairing alone, and ceRDB, respectively.
Case Study 2
Breast cancer is a common disease in women, and its incidence is still increasing25 despite great improvement in therapies and early screening.26 Estrogen receptor–positive (ER+) breast cancer and ERBB2-positive breast cancer (≈50% coexpressed with ER+ tumors) currently account for ≈75% and 15% of all breast cancer cases, respectively. The remaining 10% are so-called triple-negative breast cancers, as defined by absent expression of ER, progesterone receptor, and ERBB2 proteins.27,28 In this study, our objective is to identify genes mediated by the estrogen receptor α, ESR1, through ceRNA regulatory network in breast cancer. To this end, we selected ESR1 as the GOI and then predicted microRNAs as GTmiRs (Table 2, column 2). A total of 730 predicted ceRNAs by sequence pairing were obtained by SVMicrO at P value <0.05. Top 20 predictions were provided in Table 5. Predicted breast cancer–specific ESR1 ceRNAs by coexpression test were subsequently obtained by including TCGA gene expression of 590 breast cancer samples (described in Methods section). A total of 378 breast cancer–specific ceRNAs were obtained and top 20 are shown in Table 5.
To substantiate our finding, we examined the gene regulation networks modulated by ESR1 ceRNAs in different breast cancer subtypes. As classified by earlier studies,19 4 major subtypes, determined by molecular signatures, are luminal A (ER+, progesterone receptor–positive, Her2−), luminal B (ER+, progesterone receptor–positive, Her2+), basal-like (mostly triple-negative breast cancers), and Her2 (amplified or overexpressed ERBB2). To construct subtype-specific ESR1-mediated ceRNA networks, we prepared 4 TCGA expression data sets for the corresponding 4 subtypes, which included 93, 56, 228, and 123 samples for basal-like, Her2, luminal A, and luminal B, respectively. Considering that genes may express constantly within each subtype, we added all normal reference samples to each subtype to increase the dynamic range for correlation analysis. For each subtype, coexpression analysis was performed and integrated with sequence paring predictions (Table 5). To generate the interaction network, the process was repeated on top 10 predicted ceRNAs. Figure 4A to 4D illustrates the resulting subtype-specific ceRNA networks. Among these networks, only 2 first-layer ceRNAs (NOVA1 and CPEB3) are shared. NOVA1 has been implicated in breast cancer29 and correlated in gene expression with ESR1,30 and CPEB3, a regulator of EGFR,31 has been shown to be important in breast cancer.32 Although these 2 genes’ expression levels are mediated by ESR1 via GTmiRs in all 4 subtypes, other unique ceRNAs are also important to each subtype. For example, ceRNA MAX in basal-like regulation (Figure 4C) is an important partner of proto-oncogene Myc in driving cell proliferation in a variety of tumors. In the case of Her2 (Figure 4D), ESR1 interacts with ANK2 and then PAX2, another gene that plays a critical role in breast cancer.33
Here we presented TraceRNA, an easy-to-use Web application for predictions of ceRNAs of a GOI and their interaction network. This Web application is motivated by lack of ready-to-use tools for ceRNA predictions, and to the best of our knowledge, TraceRNA is the only Web application other than ceRDB that is specialized for ceRNA predictions. Compared with ceRDB, TraceRNA has richer functionality designed to meet different research needs. Because of the high false-positive rate and low sensitivity of existing microRNA target prediction algorithms, TraceRNA provides the users with 3 different algorithms so that they can compare or complement results as a remedy to the potentially poor predictions from a single algorithm. TraceRNA also includes the validated targets from mirTarBase, which can be selected to potentially improve the specificity of ceRNA predictions. TraceRNA also enables predictions of context-specific ceRNAs by integrating coexpression with sequence-level predictions. In the current version, 2 expression data sets from TCGA have been preloaded into the Web database. All prediction results include P values and false discovery rate as statistical significance. What is also unique about TraceRNA is its ability to construct and plot ceRNA interaction networks. This network can reveal important interactions that might not be easily perceived with the list of predicted ceRNAs. The generated network plots can be downloaded and are ready to be used for scientific publication. In contrast, ceRDB includes only 1 algorithm for microRNA target predictions and is also devoid of the aforementioned functions in TraceRNA.
Two case studies, prediction of PTEN ceRNAs and that of ESR1 ceRNAs, were presented to demonstrate the effectiveness of TraceRNA in making biologically meaningful predictions. Because both genes are important cancer-associated genes, their ceRNAs in the context of cancer were also predicted. In the case of PTEN, the GBM-specific ceRNAs were shown to be functionally more enriched than the sequence-level predictions alone, and important signaling pathways known to be related with PTEN regulation were also predicted among the most enriched pathways. When compared with ceRDB, TraceRNA predictions were functionally much more enriched, indicative of higher prediction specificity. For ESR1, unique ceRNA interaction networks for 4 breast cancer subtypes were constructed. Although ceRNAs common to 4 networks were observed, considerable differences exist among these 4 networks in ceRNAs and their interactions. Examples were provided to show possible links between the unique ceRNA interactions and the subtypes, which suggests that these differences in ceRNA interactions may well be used to explain the genomics mechanisms underlying the subtypes. If proven true, the ceRNA networks could provide an alternative to the genomics markers for disease treatment. Taken together, TraceRNA has been shown as an effective tool for context-specific ceRNA predictions and discovery of ceRNA interactions modulated by GTmiRs.
Context-specific ceRNAs are closely dependent on microRNA expressions. A predicted ceRNA by sequence pairing could not compete with GOI for binding of weakly expressed microRNAs. As a result, only highly expressed microRNAs should be considered in ceRNA predictions. Current version of TraceRNA does not yet consider microRNA expression for ceRNA predictions but displays expression values to help the user to select GTmiRs. On the contrary, mRNA expression profiles are still much more accessible than microRNA expression data; ceRNA predictions based on mRNA expression will still be of higher interest in practice. However, as microRNA profiles become increasingly available, there will be more demand to include microRNA expression in ceRNA predictions to achieve more accurate context-specific predictions. Future work should allow us to incorporate this function in TraceRNA to enable predictions in a microRNA expression–dependent fashion.
We thank the computational support from the UTSA Computational System Biology Core, funded by the National Institute on Minority Health and Health Disparities (G12MD007591) from the National Institutes of Health. TraceRNA is freely accessible at http://compgenomics.utsa.edu/cerna. As a Web application, there is no requirement for the users to access the application other than internet connection and a browser.
Sources of Funding
This work is supported in part by a National Science Foundation grant (CCF-1246073), a Qatar National Research Fund grant (09-874-3-235), and National Institute of Health grants (NIH-NCATS UL1TR000149 and U54 CA11300126; Integrative Cancer Biology Program).
↵Guest Editors for this series are David M. Herrington, MD, MHS, and Yue (Joseph) Wang, PhD.
The Data Supplement is available at http://circgenetics.ahajournals.org/lookup/suppl/doi:10.1161/CIRCGENETICS.113.000125/-/DC1.
- © 2014 American Heart Association, Inc.
- Kozomara A,
- Griffiths-Jones S
- Yue D,
- Meng J,
- Lu M,
- Chen P,
- Guo M,
- Huang Y
- Hsu SD,
- Lin FM,
- Wu WY,
- Liang C,
- Huang WC,
- Chan WL,
- et al
- Betel D,
- Wilson M,
- Gabow A,
- Marks DS,
- Sander C
- Wang X
- Maragkakis M,
- Reczko M,
- Simossis VA,
- Alexiou P,
- Papadopoulos GL,
- Dalamagas T,
- et al
- Benjamini Y,
- Hochberg Y
- Hotelling H
- Aslam JA,
- Montague M
- Shannon P,
- Markiel A,
- Ozier O,
- Baliga NS,
- Wang JT,
- Ramage D,
- et al
- Kamangar F,
- Dores GM,
- Anderson WF
- O’Brien KM,
- Cole SR,
- Tse CK,
- Perou CM,
- Carey LA,
- Foulkes WD,
- et al
- Buckanovich RJ,
- Darnell RB
- Peng SC,
- Lai YT,
- Huang HY,
- Huang HD,
- Huang YS
- Pitteri SJ,
- Amon LM,
- Busald Buson T,
- Zhang Y,
- Johnson MM,
- Chin A,
- et al