Z-scores

[Eamon Fitzgerald]

Hello,

I have a brief question about iLINCS, but first thank you for the software- it's very useful! 

My question is about running an enrichment analysis in the various databases with a list of user genes. This is very straightforward to do, but the results come in Z-scores and my question is- are these Z-scores always converted to positive? For example, using some gene lists I get high Z-scores for compounds that have opposing functions (e.g., pro-inflammatory LPS and anti-inflammatory drugs) and was wondering if the analysis returns the absolute Z-score of all significant results? If this is the case, is it possible to have the software return results predicted to both produce and reverse the inputed gene list?

Best wishes,

Eamon

[mario.me...@gmail.com]

Hi Eamon,

The enrichment results (z-scores) are based on two-tailed p-values of expression changes of submitted genes, and it is always positive. We are planning on adding directionality in a future refresh. 

At this point, my preferred way of figuring out what exactly is going on is to select top signatures (perhaps top 100) and then select "Group analysis". This will create interactive heatmaps of expression changes of queried genes and you can quickly see which enrichment is driven by upregulation and which by downregulation of queried genes. 

Best,

Mario