Annotation of unigenes
To provide putative annotations for the assembly, all of the assembled unigenes were evaluated using BLASTx and BLASTn similarity searches against the Nt, Nr, Swiss-Prot, COG and KEGG databases (Table 3 and Table S1). A total of 24,767 unigenes (18.4% of the total) were matched in the Nt databases, and 38,985 unigenes (28.9% of the total) matched in the Nr databases with an E value <1e-5. Additionally, 27,849 unigenes (20.7% of the total) can be matched to the Swiss-Prot databases with an E value <1e-10 (Table 3).
10.1371/journal.pone.0079516.t003 Table 3  Unigene annotation statistics of the C. fluminea transcriptome.   Annotation information could not be assigned for a large percentage of the sequences obtained in this study. The poor annotation efficiency could have been due to the lack of sequences in public databases for species that are phylogenetically closely related to C. fluminea [30]. Only 40 (0.03%) unigenes were matched to C. fluminea. For matches to the Bivalve class in the Nr database, the greatest number of the matched unigenes (0.4%) showed similarities with Ruditapes philippinarum followed by M. galloprovincialis (0.39%), Crassostrea gigas (0.35%), Chlamys farreri (0.21%) and Haliotis discus discus (0.17%). The low number of matches indicates a lack of bivalve data in public databases.