ToppMiR is a web-based one-stop integrated meta-analytical server to rank and analyze the significance of miRNAs with respect to user-defined biological context including an optional input of mRNA and miRNA expression profiles. ToppMiR learns intrinsic and hidden knowledge from the context by recognizing significant features of the gene sets. And the mRNA or gene ranking (target and non-target genes) is based on previously published ToppGene and ToppNet applications. Additionally, ToppMiR also ranks the microRNAs integrating the target predictions (compiled from several different prediction algorithms) and the putative targets’ relative importance. Users can optionally use expression profiles to further refine the miRNA-mRNA interactions. ToppMiR further enables extraction and export of either entire or partial networks of miRNAs, genes and annotations under analysis in a variety of formats (e.g. Cytoscape and Gephi to facilitate further analyses. Below is a schematic view of the application.
ToppMiR is designed to handle both training set-defined and -independent scenarios. The general workflow of ToppMiR is:
Below is a layered view of the general workflow.
1. Paste your miRNAs of interest and specify the identifier (required). Your choices include:
2. Paste your mRNAs of interest and specify the identifier (required). Your choices include:
3. Check the checkbox if you would like to input a proper training set of genes to be added in the analysis (optional). And paste your training set of genes in the emerged text area.
1. You may click ► to select one or more miR-mRNA sources of your preference. They are all selected by default. Your choices include:
2. Load an example set of miRs and mRNAs to learn ToppMiR.
3. Paste your training set of mRNAs if applicable and then proceed by "Next" button.
1. Resolved miRs of input displayed on the left panel. No. of putative targets in training set if applicable.
2. No. of putative targets in test set.
3. Alternative miRs suggested by ToppMiR. Check the box if you want to include the candidate in the following analysis.
4. Click "Proceed" button when you confirm the above information.
1. Putative targets in the training set if applicable.
2. Non-regulated mRNAs in the training set if applicable.
3. Overlapping mRNAs in both training and test set if applicable.
4. Putative targets in the test set.
5. Non-regulated mRNAs in the test set.
1. Check the categories you would like to perform the enrichment analysis on. A default set of categories are checked. Select the correction method you would like to apply on the annotations retrieved from the enrichment analysis on gene set. Your choices are:
2. ToppMiR is going to prioritize genes based on their connectivities to the annotations. Choose the category-wise prioritization analysis approach: InDegree or SALSA.
3. Press "Proceed" button after you determine the analysis setting.
By default, top 10 annotations of each category are displayed. Click "more annotations and details" to see all annotations of this category.
ToppMiR prioritizes miRs based on their connectivities to their target mRNAs.
1. Use only top ranked targets to prioritize miRNAs. The default is set to 40%.
2. Press "Proceed" button when ready.
ToppMiR enables extraction and export of either entire or partial networks of miRNAs, genes and annotations under analysis in a variety of formats to facilitate further analyses.
Check the miRs and putative targets you would like to be included in the network. At least one target should be checked.
Layout choices include:
Method chocies include:
Format chocies include:
Annotation p-value cutff choices include:
Annotation sources include: (training set-dependent)
Press "Export Network" to generate the network you have selected.
Different sizes of miRs and mRNA targets reflect their functional relevance:
Exported biological annotations are based on their p values, target mRNAs in the test set are highlighted.
1. If this box is checked, ToppMiR will perform Functional Enrichment Anaysis on both test set and training set. Otherwise, ToppMiR will perform ToppGene to prioritize the test mRNAs.
2. Click "Proceed" to continue the analysis.
ToppMiR ranks genes from training set based on their connectivities to the enriched biological concepts.
ToppMiR will prioritize genes of test set compared to training set from protein-protein interactions.
1. User may choose how to process the ranked training set of genes.
2. Back probability: in each iteration of random walk, the likelihood of jumping to the root set (the training set.) It is set to 0.4 by default.
2. Press "Proceed" button when ready.
Ranking of training and test sets of genes after PPI analysis.
ToppMiR performs functions enrichment analysis on test set.
1. Enhance factor: multiplication coefficient applied to a enrichment term overexpressed in both training and test sets.
2. Press "Proceed" button to perform functional analysis.
Ranking of test set of genes. Left panel: putative targets ranking; right panel: ranking of entire test set
1. If box is checked, miRs will be ranked based on their connectivities to their targets in both test and training sets; otherwise, only targets in the test sets will be taken into account.
2. Press "Proceed" button to perform functional relevance ranking of miRs.
Functional relevance ranking of putative targets and miRs.
Like training set-free scenarios, users can export networks of miRs, targets and annotations in various forms. At least one putative target from test set must be selected.
If the box in the first screenshot is unchecked, ToppMiR will perform ToppGene prioritization on test set of genes in a single step.
ToppMiR is capable of integrating expression profiles with functional relevance analysis. Check the box if you have contextual expression profiles of miRs and/or mRNAs.
1. Select if the expression profile is normalized or raw. If the expression profile contains raw data, ToppMiR will perform a log2 transformation.
2. Upload expression profiles from a single text file.
3. Load sample miR expression profiles.
4. Press "Proceed" button to process the text file.
ToppMiR will read 'Expression Level' and 'Fold Change' columns from the text file.
Functional relevance, expression level and fold change (if applicable) will be combined in Euclidean space with specified coefficient. The greater the coefficient, the more important that dimension will be.
Integrative ranking of miRs and targets. In this demo, only the integrative ranking of miRs is displayed.
Chen, J., Bardes, E., Aronow, B. and Jegga, A. (2009). "ToppGene Suite for gene list enrichment analysis and candidate gene prioritization"
Nucleic Acids Res. 37(Web Server issue): W305-W311.
Bastian, M.,Heymann, S. and Jacomy, M. (2009). "Gephi: an open source software for exploring and manipulating networks." International AAAI Conference on Weblogs and Social Media.
Sonkoly, E. and Pivarsci, A (2009). "microRNAs in Inflammation." International Reviews of Immunology.28(26): 535-561.
Shannon, P., A. Markiel, et al. (2003). "Cytoscape: a software environment for integrated models of biomolecular interaction networks." Genome Res 13(11): 2498-504.