The Analyze | Reverse Translation menu option lets you create a DNA sequence from a Protein sequence, reverse translated using a specific Genetic Code (by default, the Universal Genetic Code). The default option creates a DNA sequence with N’s and other ambiguities reflecting the degeneracy of the genetic code. This is great if you want to identify less ambiguous sections to design probes or primers and in fact MacVector will even display a list of probes with the least ambiguities.
However, MacVector also offers an optimization function if you are interested in designing a gene with codon usage optimized for expression in a particular organism.
To use this function, you do need to supply a codon usage table – a number of common tables are shipped with MacVector:
/Applications/MacVector/Codon Bias Tables/
There are four different algorithms that MacVector provides for optimizing codon usage.
Most Frequently Used Codon – this simply uses the most commonly occurring codon for each amino acid. So if, e.g. the most common Leu codon is CTC, all Leu codons will be CTC. Perhaps this is only useful if you want to design a “best guess” primer and are willing to accept a certain failure rate. If you used this to optimize expression, the host would likely run out of that tRNA and you wouldn’t see optimal expression.
Frequency Distribution – this selects a random codon for each amino acid, biased towards the most commonly used codon that encodes each amino acid. Each time you run the algorithm, a different, random set of codons will be selected. If you were to generate a new DNA over and over again, eventually this would create a collection of sequences where the average codon usage would exactly match the average for the .bias organism. But any individual reverse translation may randomly be quite different.
Probability Distribution – this is probably the most powerful setting if you are interested in expression. Similar to the Frequency Distribution, this chooses a random codon, biased towards the most frequently used codons for each amino acid. However, this version tries to ensure that the final DNA sequence has a codon usage profile as closely matching as possible to the codon usage of the selected .bias file. Again, each time you invoke the algorithm, it will produce a different sequence. But as the overall codon usage in the DNA sequence is guaranteed to be as close as possible to the codon usage in the .bias organism this should, in theory, give you the best chance of high expression. Again, you will get a different sequence each time you invoke this.
Uniform Distribution – this ignores the usage of each codon and randomly assigns an appropriate codon for each amino acid. Its similar to the default algorithm that uses ambiguities to create an “absolute” coding DNA, but here it just chooses a random codon with no regard for codon usage probability. Again, you will get a different sequence each time you invoke this.