The aim of this project was to identify and annotate the selenoproteins, their homologs and the machinery required for their synthesis encoded in Chelonoidis abingdonii’s genome. A total of 37 proteins from Homo sapiens, 1 from G. gallus (Selpb) and 5 from A. carolinensis (SBP2, SelP, SECp43, SelR2 and PSTK, which have also been studied from human proteins) have been compared with an homology-based approach with the genome of Chelonoidis abingdonii using bioinformatic programs.

From a total of 43 proteins, the following characterization has been obtained:

• Selenoproteins*: DI1, DI2, DI3, GPx1, GPx2, GPx3, GPx4, Sel15, SelH, SelI, SelK, SelM, SelS, SelN, SelP, SelR1/MsrB1, SelT, SPS2, SelW1, TR1, TR2, TR3 and SPS2.
• Cysteine-containing homologous: GPx7, GPx8, MsrA, SelR2/MsrB2, SelR3/MsrB2, SelU1, SelU2, SelU3, SelW2 and SBP2.
• Other amino acid-containing homologous: SPS1.
• Selenoprotein machinery**: eEFSec, SBP2, SPS1, SPS2, PSTK, SECp43 and SecS.
• Non-predicted selenoproteins: Selpb and SelO.
• Proteins that are not present in Chelonoidis abingdonii: SelV, GPx5 and GPx6.

*Note that for some predicted selenoproteins we have not been able to find the selenocysteine residue due to annotation or scaffold prediction mistakes.

**Some of the selenoprotein machinery proteins are also classified as selenoproteins or Cys-containing homologs.

From the obtained results we conclude that 40 out of 43 studied proteins have an homologous in Chelonoidis abingdonii, and we were able to predict 38 out of the 40 proteins that were supposed to exist evolutively.

Further research is needed in order to correctly annotate the proteins whose prediction did not start with a Met (DI1, GPx1/4/7, eEFSec, SECp43, MsrA, Sel15, SelM, SelS, SelU2, SelW1, SelW2, SPS2, TR1, TR2, TR3, SBP2, SelP, SelN, SelO, SelR2, SelR3, SelU2) or that had some gaps. For most predicted selenoproteins we could identify a valid SECIS element and find a Sec residue (except for SelK, SelO, SelW1 and SPS2). However, several times Seblastian software was not able to find them. We think that this corresponds to a false negative.

On the whole, we consider that a relatively high number of proteins have been characterised in Chelonoidis abingdonii and we believe our findings represent a subtle contribution to the current knowledge of selenoproteins.