Functional characterization for microRNAs biomarkers in lung cancer study.

• We have normalized counts in 849 microRNAs for a total of 30 individuals: 23 lung cancer patients and 7 healthy ones.
• These data were obtained after applying a primary analysis that included the evaluation of the sequencing quality, mapping and quantification of expression at the miRNA level.
• The first 23 samples correspond to microRNAs in patients with lung cancer (LUAD) and the last 7 to healthy people (CONTROL).
• In Babelomics this data was loaded. Samples were then edited to indicate the group to which each sample belongs and finally a differential expression analysis was performed. We obtained these results when comparing LUAC vs CONTROL:
  • 276 significant microRNAs down-expressed in LUAC
  • 196 significant microRNAs up-expressed in LUAC
  • All microRNAs ranked by t statistic (differential expression)

From the previous results in differential expression analysis of microRNAs, we would like to perform several functional approaches that help us understand the functions associated with these biomarkers of interest:

1. Go to mirPath webtool.
2. From All -microRNAs ranked by t statistic, we have selected 10 microRNAs more up-expressed in LUAC: file.
3. Now we would like to know functions associated to its target genes (over-representation methods). Several possibilities:
   • After uploading this file, you have to select:
     • organism: human
     • database to find information about target genes for each microRNA: microT-CDS to start
     • way to merge results: genes intersection, genes union, pathways intersection, pathways union. By default, the enrichment analysis include all target genes together. This option is ok. Run this job: KEGG analysis.
   • Are there significant KEGG pathways to explain the functional rol of target genes for these 10 selected microRNAs?
   • Go to the tab “GO analysis”. Did you detected significant GO terms?
   • Could you repeat both analysis (GO & KEGG), but previously we change the information to define target genes: now we select TargetScan in place of microT-CDS. Are there more or less genes associated to each microRNA? Why?
   • After running these jobs, do you have any differences between the first approach and the second one when using microT-CDS? Do you think the annotation genes-microRNAs is important to do this over-representation analysis?
4. A new analysis strategy for these 10 selected microRNAs:
   • Organism: human
   • Database to find information about target genes for each microRNA: microT-CDS
   • Way to merge results: pathways union. This option generates an enrichment analysis for all genes associated to a specific microRNA, and then it shows results for all microRNAs at the same by heatmaps and clustering trees.
   • Run this job: KEGG analysis.
   • Some questions:
     • Are there significant KEGG pathways?
     • Could you explain the meaning of graphical representations?

1. Go to omiRas webtool and explore options to work.

1. From miRNet you check its several resources