Comparing the Continuous Representation of Time-Series Expression Profiles to Identify Differentially Expressed Genes. Ziv Bar-Joseph, George Gerber etc.
Khalid Syed Raamesh Deshpande
Synopsis Identify differentially expressed genes between two non homogeneous time series gene expression data.
Application Cell Cycle, Immune Response, Drug effects, Infection, Response to environmental condition etc.
Nine of the cell cycle regulated genes that were identified by the algorithm as differentially expressed.
16 most significantly changed non cycling genes.
Preprocessing of Time-Series Micro Array Data
How do we know that the sampling rate was adequate (one or both of the datasets may be under-sampled)?
Interpolating and aligning data with limited number of time points can be highly susceptible to error. Cubic Spline Interpolation can also act as “smoothing” the data. The preprocessing of the data may flatten Some sudden changes in gene expression.
Methodology / Algorithm
What happens if we apply this algorithms on Datasets which have similar expression but one is time shifted (phase shift).
Let’s assume we have two time series data aligned with the same sampling points. Can we use direct pointwise comparison (or static comparison). Why or why not? Can this algorithm be extended to discover causal relationships between the genes?