Python Programming for Biology: Bioinformatics and Beyond

Figure 21.14.  Example state probabilities, as calculated by the forward-backward

algorithm. A plot of the probabilities for the buried (black) and exposed (grey) protein

amino acid states at each position of the Markov chain. Where the probabilities are

distinctly different we can be more confident about the category assignments.

1

An enzyme that cuts DNA at specific sub-sequences.

2

If a whole-genome sequence is not available this could also be done the old-fashioned

way, by hydrolysing the DNA into its component nucleotides and chemically determining

the concentration of the components.

3

 Also,  if  we  are  thinking  about  DNA  from  different  individual  organisms  which  have

slightly different sequences, compared to the sequenced genome, then the fractions would

vary by a tiny amount, though this would usually be of little concern.

4

 This  is  actually  an  axiom:  for  mutually  exclusive  or  disjoint  events  Pr(E

1

 or  E

2

)  =

Pr(E

1

) + Pr(E

2

).