If we want to change DNA code to a computer language, it will be 2 based or 4 based?

[cheetaman]

some people say it should be 3-based.

 

There are 4 DNA letters, A, T, , C. Anyway all DNA word have three letters.

[Strange]

It depends (as always with questions like this) what you want to model and why.

If you want to model at the level of bases in DNA then you will need 4 or maybe 5 values to represent them (remember, RNA uses uracil instead of thymine).

But you might need more if you want to model the artificial DNA created recently that used 6 base pairs.

If you want to model the codons, then you will 64 values. But if you were only interested in the amino acids that they encode, then you would only need 20 values.

 

[Sensei]

In computers there are needed just 0 and 1, bit, to make the all other possible combinations, after using enough of bits together.

But it's wise to have checksum, to be able detect damage in code.

Damage in DNA is inevitable, simply because of some of radioactive Carbon C-14 will decay to Nitrogen-14, sooner or later:

[math]_6^{14}C \rightarrow _7^{14}N + e^- + \bar{v}_e + 156 keV[/math]

And such data is destroyed in unpredictable way.

Edited April 14, 2018 by Sensei