CLUSTERING A SERIES OF REPLICATED POLYPLOID GENE EXPRESSION EXPERIMENTS IN MAIZE

Lingling An, Nicole C. Riddle, James A. Birchler, R. W. Doerge
2006 Conference on Applied Statistics in Agriculture  
Ploidy level is defined as the number of individual sets of chromosomes contained in a single cell. Many important crop plants, such as potato, soybean and wheat are polyploid. Although it is widely known that polyploidy is a frequent evolutionary event, it is not fully understand why polyploids have been so successful. In this work cluster analysis is employed to study gene expression changes in a maize inbred line (B73) across a range of polyploidy levels. The B73 ploidy series includes
more » ... oid, diploid, triploid and tetraploid plants and consists of biological and technical replicates as measured by microarray technology. An improved version of CORE (iCORE; improved Clustering of Repeat Expression) is presented to differentiate highly negatively correlated genes while taking advantage of the additional information that is provided by replication. The error information from the replicate experiments is utilized to cluster gene expression for both simulated and real ploidy-series data. Simulation results indicate that iCORE leads to an improvement in accuracy over both CORE and hierarchical clustering based on average gene expression only. When applied to the maize ploidy series, the iCORE results provide information that may aid in understanding of the effect of gene dose on gene expression in a ploidy series.
doi:10.4148/2475-7772.1120 fatcat:nrwhl2wdvvbvtcwjlx6prnk2pu