The DNA Pac-Man game (https://github.com/HussainAther/dnapacman) can represent how protein sequences are generated (using Hidden Markov Models). We can draw an analogy of how HMMs work in the context of generating protein sequences using the Pac-Man. The next token/letter you eat is the next letter in the sequence of generating a sequence of protein amino acids.
Essentially, if we organize regions of the Pac-Man board representing different hidden states in a Markov Model, then the next letter that Pac-Man eats can represent the next state an HMM selects. We can change the probabilities a certain letter may appear and, when Pac-Man enters the hidden state, then the probabilities would change.
We can observe how the probability for different states changes as people play based on which letter they choose next. We can compare the HMM of Pac-Man to the HMMs in modeling eukaryotic genes following the methods of this manuscript “Hidden Markov Models and their Applications in Biological Sequence Analysis”: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2766791/
Full post here: https://github.com/HussainAther/DNAPacManHMM