CONCLUSIONS

A summary of the project will be presented in order to have an overall view. Selenocysteine is considered to be the 21st amino acid. Hence, the proteines that include this amino acid within their sequence are called selenoproteins. They are mainly known because of their important roles in oxidative damage protection.

Despite its importance, selenoproteomes are not properly described in many several organisms, and this could lead to misinterpretations since the codon coding for selenocysteines is UGA, the same as for the STOP codon. Thus, the aim of this project was to annotate the selenoproteome of an organism, Urocitellus parryii, which had not been characterised before.

Our results altogether are synthesized here:

- 21 Selenoproteins were found: DIO1, DIO2, DIO3, GPx1, GPx2, GPx3, GPx4, GPx5, GPx6, Sel15, SelI, SelK, SelM, SelN, SelO, SelP, SelR1, SPS2, TR1, TR2, TR3.

- 8 Cys-containing homologs: GPx5, GPx7, GPx8, MrsA, SelR3, SelS, SelU1, SelW2.

- 2 selenium molecular machinery: SBP2, eEFsec.

- 1 other amino acid containing homolog: SPS1.

- 5 unclassified: SelH, SelR2, SelU2, SelU3, SelW1.

Note that the ‘Unclassified’ section stands for those proteins which, according to the alignment results from T-coffee, contain a selenocysteine, but do not have any valid SECIS prediction. We considered useful to classify them this way because even if the prediction showed a possible presence of a determined selenoprotein, the apparent lack of SECIS elements may lead to the interpretation of the codon as a stop codon, which is not the case since more amino acids are codified after the predicted ‘selenocysteine’. Hence, there are two possible explanations for these cases: a) Neither seblastian nor SECISearch3 predicted a SECIS element, but maybe it is present and could be found using other tools, b) there has been an annotation error, placing a selenocysteine where it does not correspond.

Moreover, it is worth mentioning the lack of SelV within Urocitellus parryii’s genome. Besides, SelW1 was supposed to be a selenoprotein. However, the first part of the protein was missing (no more scaffolds appeared in the tblastn) and it is where the selenocysteine is found in the human’s genome. Hence, we could not predict whether this is a selenoprotein in U. parryii.

We considered also practical to distinguish between the sequences that were partially and fully predicted, even though almost every prediction was complete.

- Fully predicted: DIO1, DIO2, DIO3, GPx1, GPx2, GPx3, GPx4, GPx5, GPx6, GPx7, GPx8, MrsA, SEL15, SELH, SelI, SelK, SelM, SelN, SelO, SelP, SelR1, SelR3, SelS. SelU1, SelU2, SelU3, SelW2, SPS1, SPS2, TR1, TR2, SBP2,eEFsec

- Partially predicted: SelR2,TR3.

Among the limitations that we found, it is worth to say that the exonerate that we did, predicted different genes that did not have biological reliance. Moreover, the genome of our animal may not be perfectly annotated as in some cases we found out that it lacks a methionine in the beginning of the proteins.

In conclusion, the prediction of the selenoproteome of this organism may be useful and practical for further studies. However, some limitations should be overcome in order to have even better results and more research is needed in this field.