Searching for Genes

Table: Selectors for the NucPepMatch structured type.

1.1 

Table: Selectors for the NucPepMatch structured type.

Selector	Description
Sim	Similarity score for the match.
NucOffset	Offset of the nucleotide sequence in NucDB.
PepOffset	Offset of the nucleotide sequence in PepDB.
NucLength	Length of the nucleotide sequence.
PepLength	Length of the peptide sequence.
PamNumber	The estimate PAM distance for the match.
PamVariance	The estimated PAM variance for the match.
IntronScoring	The function for scoring introns.
NucGaps	Gaps in the nucleotide sequence from the alignment.
PepGaps	Gaps in the peptide sequence from the alignment.
Introns	The list of suspected introns.

 

Assume we have just sequenced the following two chunks of RNA from some prokaryote.

> chunk1 := 
>   'UAUUGCAUAUCGGCAUUGGCGGUUCUGCUCUCGGCCCCAAAUUGCUGAUUGAUGCCCUGACU'.
>   'CGUGAAUCUGGCCGCUAUGAUGUUGCGGUUGUUUCCAAUGUGACGGUCAGGCUCUGGAAGAA'.
>   'GUCUUCAAAAAAUUCAAUCCGCAUAAAACCCUGAUUGCGGUUGCUUCCAAGACCUUCACCAC'.
>   'CGCAGAAACCAUGCGAACGCAGAAUCCGCGAUGGAAUGGAUGAAGAAGCAUGGCGUCGAAGA'.
>   'UCCUCAGGGUCGCAUGAUUGCCCUUACCGCAAUCCGGCGAAGGCUUCUGAAAUGGGUAUUGAU':
> chunk2 :=
>   'GAAACAUCGGCGGCCGCUAUUCUCUGUGGUCUUCCAUCGGUUUUCCAGCGGCGCUGGCUCUU'.
>   'GGAUGGGAAGGCUUCCAGCCAGCUUCUCGAAGGUGGCGCGGCUAUGGAUCGC':

We assume the 311 and 114 base chunks to be part of the same mRNA molecule, but we do not know

• whether both chunks belong to a single mRNA molecule (they obviously do not overlap),
• whether chunk1 follows chunk2 or vice versa,
• how many bases are missing between the two chunks,
• what peptide(s) the chunks might code for.

We are going to answer these questions as follows:

1.

Search the SwissProt protein database for candidate proteins being reasonably similar to one derived from some encoding of our chunks.

2.

Once candidates in SwissProt have been detected, find the proper alignment with the original chunks using direct nucleotide versus peptide dynamic programming.