Meta Analysis - Death List

The HTS-DB currently holds 13 human cell line genome-wide screens and 2 drosophila genome-wide screens. 12 of the 13 human genome-wide screens have cell viability data available, this includes 2 screens of lung-cancer PC9 cell lines, 3 screens of kidney cancer lines (RCC4, A498 and 786-O lines) and 2 screens of breast cell lines (MCF-10A and MCF-7).

In order to rank genes in terms of their knockdown effect across all genome-wide screens in our database we employed the redundant-siRNA activity (RSA) measure. This method was used to rank siRNA pools from the entire genome collection according to normalised experimental effect on cell viability. A p-value for each single pool based on whether the scores for that pool are distributed significantly higher in the rankings than would be expected by chance could then be calculated.

This accumulated cell viability data is useful when hit calling as it provides a landscape of frequently occurring activities against which we can compare every new dataset. For example, several of our screens have attempted to identify genes that when knocked down have a differential effect in two cell lines or conditions. Information from our meta-analysis has proved useful to classify those hits that are frequently involved in loss of cell viability as opposed to those hits where loss of viability is more significant for the experimental set up and cell line. It has also been useful to highlight unusual differences in phenotypes between groups of cell lines or conditions.

Click [here] to access the supplementary table containing all cell line Z-scores and RSA analysis.

String database representation of interactions between the top 250 genes that are most commonly associated with loss of cell viability when knocked down in 12 genome-wide screens of 11 different cell lines.