Gene expression imputation techniques for robust post genomic knowledge discovery

Microarrays measure expression patterns of thousands of genes at a time, under same or diverse conditions, to facilitate faster analysis of biological processes. This gene expression data is being widely used for diagnosis, prognosis and tailored drug discovery. Microarray data, however, commonly contains missing values, which can have high impact on subsequent biological knowledge discovery methods. This has been catalyst for the manifest of different imputation algorithms, including Collateral Missing Value Estimation (CMVE), Bayesian Principal Component Analysis (BPCA), Least Square Impute (LSImpute), Local Least Square Impute (LLSImpute) and K-Nearest Neighbour (KNN). This Chapter investigates the impact of missing values on post genomic knowledge discovery methods like, Gene Selection and Gene Regulatory Network (GRN) reconstruction. A framework for robust subsequent biological knowledge inference has been proposed which has shown significant improvements in the outcomes of Gene Selection and GRN reconstruction methods