Recently large scale transcriptomic and proteomic datasets for human cells have become available. A striking finding from these studies is that the level of an mRNA typically predicts no more than 40% of the abundance of protein. This is largely due to regulation of translation. Interestingly, our analysis and modelling of this data defines a large and distinct group of mRNAs where there is a good correlation between mRNA and protein levels across cells. For these genes expression is canonical - mRNA levels are good predictors of protein levels. On the other hand, there are a large number of proteins for which the translation efficiency varies greatly between cells. We analysed this data to examine hypotheses of the causes of significant deviations in different cell lines. We show indications of translational control for many genes that have not yet been tested or reported. We aim to identify all the RNA elements in mRNAs that contribute to this regulation. To do this we have developed curated sets of cis-regulatory RNA elements (CisRegRNA) and novel bioinformatic methodologies to detect similar elements using support vector machines (CisRNA-SVM) and tools to analyse mRNA data (scan for Motifs). Our methodology outperforms previous methods and has discovered new structural classes of candidate regulatory elements. We are also using high throughput proteomics (pSILAC) in parallel with mRNA profiling (RNA-Seq) to identify cis regulatory elements that mediate translational control in cancer cell lines. These experimental results are being combined with the bioinformatic analysis to define the functions of novel regulatory elements.