Signals for Gene Finding

There are many signals associated with genes, each of which suggests but does not prove the existence of a gene.
Gene finding programs put together evidence from these various types of signals to suggest possible locations of genes in large unannotated sequences.
Examples of signals which can be modeled with weight matrices (in order):
- Transcription factor binding sites.
- Cap site.
- Start codon.
- Splice sites.
- Stop codon.
- Polyadenylation site.
Another important source of information for gene finding is the frequency of various codons in coding and non-coding regions. For example:
- The stop codons, TAA, TAG and TGA occur infrequently in coding regions.
- Codon's appear with different frequencies depending upon their neighbors and the local secondary structure.
The frequencies of codons in coding and non-coding regions are accounted for in gene finding programs by using a log likelihood ratio similar to a weight matrix.