New method to identify to find Reproducible Drug-resistance-related Desregulated Genes

In microarray experiments for comparing gene expression profiles of two groups of cell lines, one of the most popular experimental designs is to measure only a few (e.g., two or three) technical replicates and detect differentially expressed genes by FC with a cutoff value. Apparently, genes with high expression levels in one group are hardly to be identified to be desregulated and genes with low expression levels tend to have large FC introduced by random measurement fluctuations. Besides, the cutoff value is usually arbitrarily. Therefore, it is necessary to evaluate the character of methods that commonly used in small cell line datasets and develop some new methods which are suitable for these datasets. PD algorithm proposed in this paper can identify significantly reproducible DE genes with a stringent FDR on 4 small resistant-sensitive cell lines datasets and significantly enriched in many biological pathways known to be related with drug resistance of the corresponding drugs. SAM and RP could identify DE genes in datasets at least three pairs of samples and they were all biased to find genes with low expression levels. Besides, SAM or RP might miss many drug resistance-related biological pathways which could be identified by PD. Therefore, the PD algorithm is an effective supplementary method to SAM and RP. Refer to this link:http://www.cusabio.com/Recombinant-Protein/Recombinant-Escherichia-coli-strain-K12-Signal-peptidase-I-11089634.html