The lethal screen

To comprehensively investigate the role of microRNAs for the survival of cancer cells, we employed high-throughput functional genomics screening technology. For each cancer type, we selected a panel of commonly used cell lines and transfected them with miRIDIAN microRNA mimic or inhibitor libraries (Dharmacon; version 16.0). To select mimics/inhibitors that effectively kill cells or inhibit their proliferation, we used highly sensitive luminescent cell viability assay, which measures cellular ATP content (CellTiter-Glo; Promega). For analysis, raw luminescence values were transformed to normalised viability values, which compare to the wells treated with non-targeting siRNAs or microRNAs. When querying the database, you can adjust this parameter to search for the most effective microRNA mimics/inhibitors across different cell lines.

The database

The database was constructed from many different types of data and has been designed to facilitate the investigation of microRNAs, their predicted targets and the pathways in which the targets reside. This is achieved by focusing the search on the microRNAs and predicted targets most relevant to the system. This approach relies on a number of filtering steps which utilise the various data types in the database highlighted in the components image below and summarised as follows:

  1. A search of the lethal screen data identifies microRNAs that are lethal in a specific cell line or cell lines.
  2. MicroRNAs can be removed from the search based on their association with an increased or decreased proliferation of cardiomyocytes (Eulalio et al, 2012).
  3. A choice of five separate microRNA target databases can be selected and used to identify predicted targets. Links to these are available below in the Metric table.
  4. The list of targets can be refined by restricting targets to those expressed in a specific cell line.
  5. Targets can be refined further by limiting to targets with known essential behaviour in a specific cell line.

The final set of microRNAs and targets can then be analysed using both custom and public analysis tools. These include custom KEGG enrichment, microRNA-target networks, Reactome, STRING DB, David and EnrichR

Metrics:

Mature microRNAs (mirBase v21)2,588
Unique microRNA targets (GenCode v21)23,506
Total microRNA -> targets5,359,163
- TargetScan v6.2 (June 2012)3,254,995
- miRDB (April 2013)222,190
- miRTarBase v4.5 (Nov 2013)19,419
- Diana microT-CDS v5 (Jul 2013)1,396,442
- StarBase v2.0 (Sept 2013)466,117
KEGG pathways (Nov 2014)285
RNA-Seq expression data (Klijn et al, 2014)33,981 genes (8 cell lines)
shRNA dropout signatures (Koh et al, 2012)15,119 genes (4 cell lines)

Components:


Contact: a.swarbrick@garvan.org.au