Mechanistic Parameterization of the Kinomic Signal in Peptide Arrays

J Proteomics Bioinform. 2016 May;9(5):151-157. doi: 10.4172/jpb.1000401. Epub 2016 May 24.


Kinases play a role in every cellular process involved in tumorigenesis ranging from proliferation, migration, and protein synthesis to DNA repair. While genetic sequencing has identified most kinases in the human genome, it does not describe the 'kinome' at the level of activity of kinases against their substrate targets. An attempt to address that limitation and give researchers a more direct view of cellular kinase activity is found in the PamGene PamChip® system, which records and compares the phosphorylation of 144 tyrosine or serine/threonine peptides as they are phosphorylated by cellular kinases. Accordingly, the kinetics of this time dependent kinomic signal needs to be well understood in order to transduce a parameter set into an accurate and meaningful mathematical model. Here we report the analysis and mathematical modeling of kinomic time series, which achieves a more accurate description of the accumulation of phosphorylated product than the current model, which assumes first order enzyme-substrate kinetics. Reproducibility of the proposed solution was of particular attention. Specifically, the non-linear parameterization procedure is delivered as a public open source web application where kinomic time series can be accurately decomposed into the model's two parameter values measuring phosphorylation rate and capacity. The ability to deliver model parameterization entirely as a client side web application is an important result on its own given increasing scientific preoccupation with reproducibility. There is also no need for a potentially transitory and opaque server-side component maintained by the authors, nor of exchanging potentially sensitive data as part of the model parameterization process since the code is transferred to the browser client where it can be inspected and executed.

Keywords: JavaScript; Kinomic peptide array; Kinomics; Non-linear regression; PamGene.