GENE D

      It is in forward strand and predicted by the two programs. Geneid precists one initial, three internals and a terminal exon, while genscan predicts one initial, five internals amd a terminal exon. Genscan also predicts a promotor and a polyA, which we won't use.

      There is just one human EST in the region limitated by the predictions, but it doesn't reinforce any exon. Despite this, it has a length of 800 nucleotides, from which a 704 nucleotide region match with a socre of 99%. As it doesn't reinforce any prediction and it has 100 unaligned nuclotides, we don't consider it much informative. What we do find reinforcing our data is mouse ESTs.

      With the gene from the genscan we obtain as a result an unknown protein (protein XM_042858 for MGC:15763) with a similarity of 73% in 204 of the 328 aminoacides predicted. With the gene from the geneid we obtain a 72% of similarity with the same unknown protein, matching 213 of 222 predicted aminoacids. We will analyze every exon, but in this case, as we don't have any human EST reinforcig, we will mostly trust the genepredictions to delimitate exactly the exons.

• Initial: The initial exons predicted by genscan and geneid are exactly the same and matche very well with the one predicted doing a tblastn with the homologous protein.
• Internal:
  • Also from the homologous protein tblastn we obtain an internal exon wich isn't predicted by the geneprediction programs, but which is suported by a mouse EST.
  • The first internal exon predicted by genscan is not suported by anything else.
  • The next genscan internal exon matches 100% with 35 aminoacids of the human homologous protein. It is also reinforced by a mouse EST.
  • The next genscan internal exon is not suported by anything else.
  • The first geneid internal exon matches with a region of the human homologous protein. It is also reinforced by a mouse EST.
  • The next internal exon is the same for genscan and geneid predictions, and matches with the human homologous protein. It is also reinforced by two overlapped mouse EST.
  • The next exon is predicted as internal by genscan and geneid, though there are slight nucleotide variations. It matches with a region of the human homologous protein and is reinforced by a mouse EST.
• Terminal: The two terminal exons predicted by genscan and geneid are no supported by anything else.

  But the last exon predicted by the tblastn with the homologous protein is supported by a mouse EST. We consider this exon as good, because we search it manually in the geneid predictions. We search for a terminal exon coinciding with the one predicted with the homologous protein. We find one with good score and matching, so we select it.

      This can be visulized in GhostView format.

      Concerning the promotor sites, we do the same as in the previous case and obtain this view of the possible promotors.