This gene is in forward strand and is only predicted by genscan. This program predicted two exons (one initial and one terminal) and a polyA and a promoter sequence. We don't much trust these last two sequences because this program tends to provide them always, even if their score is very low. It isn't reinforced by any human or mouse ESTs neither.
By searching homologous human proteins, we obtain as the best matching an hypothetical human protein (XP_105662) which has an 82% of identity with 29 aminoacides from the 72 predicted. This is due to the fact that, as we can see by running a tblastn (the homologous protein versus our contig nucleotid sequence), only the terminal exon matches with the protein.
To try to reinforce the results, we search homologous proteins in mouse, but the only output we get has only 33% of homology to our sequence.
The fact that we don't find any prediction in geneid, no ESTs reinforcing the sequence and no significant homology in proteins prevents us from affirming that this is an actual gene. However, we could guess the terminal exon is correctly predicted (because of the 82% of similarity to the protein XP_105662), but as this is also an hypothetical protein we can neither asure this.
