FAQ

 
 

Color scale draws overexpression (above mean or median) in red and underexpression in blue. If data are not normalized, it might happen that most of the elements are above/below average, so the visualization will mostly use just one side of the scale. You can press ‘s’ to draw the exact scale used, skewed by the data


Make sure your data are not skewed. It is quite usual in expression experiments to get distributions with high overexpression outliers.

On these cases, log-normalization usually moderates the effect on scale of outliers.


Quite frequently, even without large outliers, expression distributions are moderately skewed towards overexpression, on these cases you can press ‘m’ to switch to quantile scale, but you must always have present the statistical properties of your data and that it affects to the color scale.

 

My visualization looks very red/blue

UPDATE: Yes, Now OBO format custom ontologies can be loaded into Voronto. You will also need a GAF format file for gene annotations on the ontology

Can I use my custom OBO ontology?

Reactome or GO are examples of deep hierarchies. In the case one of your terms of interest is quite small on the current view, try mousing over it or one of its parents, and then press ‘enter’ and the visualization will expand on that term. To go back to the general view, press ‘supr’.


Sometimes you might not see a term that you know that is in the ontology. Maybe it is too small or, if in GO, is on a level that is not yet visualized (GO is so large that only 2 levels are visualized). You can always search for it by name by pressing ‘f’. All the terms with the searched text, or that contains a term with the searched text will be highlighted. You can also search for gene names.


You might also want to check our advanced interaction walkthrough video


UPDATE: For GO, now there’s the option to start the visualization on a term different from the root node

It is hard to inspect deep terms

The number and deepness of ontology terms can make labels unreadable. We have improved the label positioning and much more labels can be now visualized in a readable way. However, several others are not, and they are hidden. You can still check them by hovering over the term with the mouse, or you can expand the visualization by hovering over it (or its parent) and pressing ‘enter’

Why I cannot see all the labels?

Which are the supported ontology versions?

Voronto currently works with the following versions:


  1. • GO (BP, CC, MF): OBO v1.2 (last update 20/03/2013)

  2. • GO (annotations): GOA files from EBI (21/03/2013)


  1. • KEGG (orthology): ko00001.keg from KEGG webpage (17/01/2013)

  2. • KEGG (annotations): retrieved by KEGG programmatic services (23/01/2013)


  1. • REACTOME (hierarchy): biopax3 owl files (02/07/2012)

  2. • REACTOME (annotations): BiomaRt BioConductor annotation package (same date)


These versions are updated periodically. If you want to use a more up-to-date version, please tell us (rodri AT usal DOT es) and we will do a new update.

KEGG colored pathways show green and/or white background components

Green background components (with black foregrounds) are the way KEGG tells that the component has no genetic elements on your expression data.


White background components (with black foregrounds) are the way KEGG tells that the component is on the general pathway but not for this organism.

Note this is different of the components colored by Voronto, which might have white (average expression) backgrounds, but always have dark green foregrounds.

Text/expression search highlights some terms in dark green and others in bright green, why?

Bright green means that the term directly contains the searched string/expression pattern on its name/profile.


Dark green means that the term does not fit the search query, but contains subterms that coincide with the search string or the expression pattern,