Details
Description
There's a subtle problem in the 'show cross-refs->Protein' workflow for a nucleotide alignment means that the CDS view doesn't include gaps inserted into the protein alignment due to misaligned codons in the original nucleotide alignment.
e.g.
>nucA
AATGAG
>nucB
-ATG
produces a protein alignment:
>pA
N-A
>pB
-M-
When the protein alignment is shown (via show crossrefs->Uniprot for the case of Ensembl: ENSG00000113924), its corresponding CDS alignment is just the spliced exons from the original Ensembl locus view. Inserts reflecting the final aligned peptides need to be propagated back to the CDS prior to display.
But the CDS reconstruction for this alignment should be:
>nucA
AAT---GAG
>nucB
---ATG---
e.g.
>nucA
AATGAG
>nucB
-ATG
produces a protein alignment:
>pA
N-A
>pB
-M-
When the protein alignment is shown (via show crossrefs->Uniprot for the case of Ensembl: ENSG00000113924), its corresponding CDS alignment is just the spliced exons from the original Ensembl locus view. Inserts reflecting the final aligned peptides need to be propagated back to the CDS prior to display.
But the CDS reconstruction for this alignment should be:
>nucA
AAT---GAG
>nucB
---ATG---