miRTrail


Tutorial

To start a new miRTrail analysis select "Home" from the menu above.

Sample datasets for miRNA and mRNA can be downloaded at the Upload-dialog under "example file".

First, the miRNA and mRNA informations must be uploaded. At the moment, miRTrail supports miRBase identifiers for miRNAs and GeneSymbol identifiers for the mRNAs. Each identifier representing a deregulated miRNA or mRNA/gene must e followed by a tab-stop and a "1" indicating an up-regulation or a "-1" indicating a down-regulation.
miRBase ID hsa-let-7a<TAB>1
hsa-miR-1<TAB>-1
hsa-miR-135b<TAB>-1
GeneSymbol HOX5A<TAB>1
TP53<TAB>-1
C11orf58<TAB>1
Upload
The next step involves the specification of a threshold for the miRNA - target predictions. A default of 0.01 is provided. Setting a smaller threshold results in fewer putative interactions. Actually, these informations are extracted from the microCosm targets service. Options

miRTrail will now perform the underlying computations and automatically present the results as soon as they are available.

Depending on the options specified (specificity of prediction threshold, all available GT-analyses etc.) it will take from 20 seconds to 30 minutes until the results are available.

The results will then be available at the link presented on the Process-page.
Process

The results will then be available for 15 days after computation. In case you desire a longer availability of your results, please feel free to contact the webmaster and, including the links of the results page(s), ask if this could be activated.

Below you will find the explanation of the results shown on the results page as well as an overview of the performed analyses.

Results - Overview Intially, a short summary of the submitted data is presented as well as an indication of the number of predicted target mRNAs of the deregulated miRNAs.
Results - Contingency analysis

The first presented analysis is an analysis of dependance/independance, based on a contingency-table and a Chi-squared test. For the specified prediction threshold, the table contains the numbers of predictions where e.g. the miRNA and the putative target mRNA are both upregulated.

Using a Chi-squared distribution with the appropriate degrees of freedom, the p-value is calculated.

Results - Overlap analysis

Next is an analysis of the statistical significance of the found overlapbetween the predicted targets of the deregulated miRNAs and the actually deregulated mRNAs.

Here, the hypergeometric distribution is used to calculate the corresponding p-value.

Results - ORA

Here, the results of the individual ORA (Overrepresentation analysis) are presented. Depending on your previous choice, either all in GeneTrail available analyses are carried out or simply the KEGG-analysis, which is the default.

Based on the three individual ORA, miRTrail checks which of the significant categories are found in more than one of those analyses, indicating an overlap between the respective category-sets.

Results - Network analysis

This part of miRTrail allows for a manual inspection of the network that arises from the submitted information and selection of individual miRNAs. Again, an ORA is available for the resulting deregulated target mRNAs.

The miRNAs of interest can be individually selected and the respective deregulated target mRNAs are listed. The miRNAs are ranked based on the number of deregulated targets and the top 5% are preselected.

Or, based on a miRNA-family enrichment analysis, the network may be constructed based on this information. Ranking of the families is done by the adjusted p-value. All families showing a significant enrichment for alpha < 0.05 are preselected.





ZBI email to webmaster
ZBI
Valid HTML 4.01 Transitional