Figure 3 depicts the four K means clustering maps that collectively make up Cluster D, while Table 4 describes transcript identity and number. The sequenced transcripts from Cluster D consist of 58% cuticle proteins, gastrolith protein and vermiform cuticle protein each constitute 2%, while 38% research use only were unable to be annotated. The group of transcripts represented in Cluster E dis play expression profiles which are relatively low during the moult, post moult and intermoult stages, then increase in the early pre moult stage and remain high in late pre moult. This group is depicted in Figure 3 where two clusters were deemed to have similar expression profiles and hence collectively termed Cluster E. Table 5 describes the composition of Clusters E1 and E2.

In this cluster 31% of sequenced transcripts are fatty acid binding proteins, carcinin and C type lectin recep tor each make up 25%, 13% are vermiform cuticle pro teins and 6% of the sequenced transcript population is clotting protein. Cluster F is depicted in Figure 3 and features tran scripts whose expression is highest in the moult and post moult stages, decreases substantially in the inter moult and early pre moult stages then begins to increase again in late pre moult in preparation for ecdy sis. Table 6 provides a description of the identity and number of the total transcript population for Cluster F. A high proportion of the sequenced cDNAs are mannose binding proteins, 23% are cuticle proteins, 8% represent myosin, while 31% remain unannotated.

The expression profiles of the transcripts comprising Cluster G show relatively low expression levels in the moult stage, an increase to peak levels in the post moult and intermoult stages, then a dramatic decrease in the early pre moult stage which begins to increase again in late pre moult. The gene expression pattern for Cluster G is presented in Figure 3 Table 7 describes the identity and number of the transcripts assigned to Cluster G, of which the sequenced transcripts comprise of 50% cuticle proteins and 50% unannotated sequences. Figure 4 presents a summary of all the expression pro files described above and allows a comparison of cluster profiles. A peak in down regulation in the early pre moult cycle can be seen in clusters D, F and G, while a peak in up regulation in this moult stage is observed in clusters A, B and E.

Discussion A holistic approach to gene expression profiling was employed in order to gain a greater understanding of the molecular events associated with the crustacean moulting process. A P. pelagicus moult cycle specific cDNA microarray, containing sequences from 5000 cDNA clones derived from whole crabs in addition to individual organs such as the brain, eyestalk, Carfilzomib MO and Y organ from all moult cycle stages, was developed for this study.