This work is aimed to algorithmically (in silico) reproduce the sequence of DNA from protein sequence, i.e. solve
protein back-translation problem or reverse translation problem. Presented solution produces a result using
hidden Markov models without usage of codon frequency data obtained in other (previous) analysis. Based on
biological foundations of protein biosynthesis the Markov model is constructed, then trained with the support
of the Viterbi algorithm, and used to estimate the most likely coding sequence. An application that implements
the above algorithm is described, the results obtained by this application are studied, and comparison to other
solutions of the protein back-translation (reverse translation) problem is presented.
© (2012) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Tomasz Kaczynski and Robert Nowak
Reverse translations of gene-coding DNA sequences using hidden Markov models
", Proc. SPIE 8454, Photonics Applications in Astronomy, Communications, Industry, and High-Energy Physics Experiments 2012, 84541P (October 15, 2012); doi:10.1117/12.2000163; http://dx.doi.org/10.1117/12.2000163