A neural network (NN) was trained on amino and nucleic acid sequences to test the NN’s ability to predict the correct codon given only an amino acid sequence. Different network configurations were used with varying numbers of input neurons that represented amino acids and a constant representation for the nucleic acid. A multi-layer backpropagation network of one hidden layer with 5 to 9 neurons was used. In the best-trained network, 93% of the overall bases, 85% of the degenerate bases, and 100% of the fixed bases were correctly predicted. The training set was composed of up to 60 human sequences in a window of up to 25 codons at the coding sequence start site. Different input configurations for amino acid representations were designed and evaluated for usage in a large scale NN. This genetic data analysis effort will assist in understanding human gene structure. Benefits include computational tools that could predict more reliably the backtranslation of amino acid sequences useful for Degenerate PCR cloning, and may assist the identification of human gene coding sequences (CDS) from open reading frames in DNA databases.
| Identifer | oai:union.ndltd.org:auctr.edu/oai:digitalcommons.auctr.edu:dissertations-2188 |
| Date | 01 July 1998 |
| Creators | White, Gilbert F. |
| Publisher | DigitalCommons@Robert W. Woodruff Library, Atlanta University Center |
| Source Sets | Atlanta University Center |
| Detected Language | English |
| Type | text |
| Format | application/pdf |
| Source | ETD Collection for Robert W. Woodruff Library, Atlanta University Center |
Page generated in 0.0016 seconds