Research » Biostatistics

Statistical techniques are useful in proteomic discovery and biomarker validation and qualification experiments. Useful methods include:Statistical techniques are useful in proteomic discovery and biomarker validation and qualification experiments. Useful methods include:

Design

Experimental design and sample size determination are critical to successful investigations. Inferences can be impossible in poor designs, where potentially substantial biases can preclude meaningful biological conclusions. The demonstration of the risk of these problems and determining novel methods to overcome them is an active area of research. 

Analysis

Analysis of biomarker discovery, validation and qualification experiments is increasingly defined by guidelines such as ReMARK. It is important to assess both statistical and clinical significance at all stages of investigations which provides challenges in analysis and interpretation. For this purpose evolving bespoke routines are available to undertake such analysis efficiently, using scripts developed for use in the R environment for statistical computing.

 

Pairwise same: same comparisons of LC-MS/MS label-free proteome of clear-cell renal cell carcinoma formalin-fixed paraffin embedded tissue for 16 patient samples informs sample size by quantifying normal biological variation. Red line shows 95th percentile quantile across abundance range (2.5 to 3-fold) indicating minimum useful difference to detect. 

 

 

Risk assessment plot for ACY-1 when added to reference multivariable model for predicting delayed graft function (DGF) in renal transplant patients. Curves are for novel predictive model (including ACY-1) and the null reference model for each of sensitivity and 1-specificity at each calculated risk for those with an event (DGF) and without an event. The area between the curves is the integrated discrimination index. 

 

 

A biomarker index score composed of four MS-discovered protein biomarkers can predict treatment benefit from standard therapy (CONTROL) or standard therapy with bevacizumab (TREATMENT) in the ICON7 trial. Kaplan-Meier estimates of survival functions for biomarker score low and high separated by treatment arm show the benefit of bevacizumab when SCORE=3 and the lack of benefit when SCORE

 

Privacy Notice | Page generation time: 0.3281