RNA modeling with Matlab - Chemistry 694 - Summer 2007
The links below are outdated. Download the zip file above and use
the instructions there.
Note: some browsers display the Matlab program files below as if they had
no carriage returns; in this case, either save the page to view it locally
or use View, Page Source.
Independent bases
-
show(S,N) uses letters N to show states S
-
rando(x) returns a random integer distributed according
to probability vector x
-
rna_iid(n) generates n independent, identically
distributed RNA bases
Markov chains
-
rna_mar(n) generates n RNA bases with Markov dependence
-
transition_count(x,N) makes a table of
transitions of x and labels the states using letters in the string N
Hidden Markov models
-
die_mar alternates between fair and loaded die
rolls
-
hmmsim(mu,A,n) generates n states of a Markov chain
with initial distribution mu and transition matrix A.
-
obssim(Pi,E) generates emitted states based on hidden
sequence Pi using emission probabilities E.
-
viterbi(mu,A,E,X) uses the Viterbi algorithm to
find the most likely hidden state sequence associated with the data in
matrix X, where mu is the initial distribution of hidden states, A is the
transition matrix, and E is the matrix of emission probabilities.
-
dice(n,r) alternates between fair and loaded dice,
runs r simulations of length n each, and estimates the hidden sequence.
The actual hidden sequence is shown first, then the observed sequences,
then the estimated sequence, with errors marked below.
-
dice(78,1) output
-
dice(78,2) output
-
dice(78,3) output
-
dice(78,5) output
-
dice(78,10) output
-
dice(78,20) output
Stochastic grammars