More errata in Eidhammer, Jonassen, and Taylor: Protein Bioinformatics

This list contains errata not mentioned on the book website:
http://www.ii.uib.no/proteinbioinformatics/errata/

----------------------------------------------------------------------------

Chapter 1

p 7, example: Wrong spacing in alignments

p 14, line 7: Figure 1.1 -> Table 1.1

p 16, top:
"This constraint also tends to prefer one longer gap over several
neighbouring short ones"
This is not true if equality holds, i.e., for linear gap penalty.

Chapter 3

p 53, 1st line: middle value -> mean value
p 55, 3.3.1 title: "The P value has an extreme value distribution"
      Should be the S_M score (the P value always has a uniform
      distribution)

Chapter 4

p 83: The formula for L_{v,w} is not correct.
p 98, exercise 7: An ultrametric tree is additive by definition (p 78).

Chapter 5

p 119, Eq. (5.8): argument of ln should be the whole fraction (h_ab/p_a p_b)

Chapter 7

p 146, Figure 7.1: "with probability 1/2 each" is true only in one
point of the plot.

Chapter 9:

p 195 after Eq. (9.3): common origin -> common centroid

Chapter 15:
p 302, just after Eq. (15.1): buried states -> burial states

Appendix A

p 315: faculty -> factorial


Clarifying remarks:

Chapter 4:

Section 4.3:
"INS1(Mouse) and INS2(Rat) are only homologs"
This definition is controversial.
Some people would classify them as paralogs because they diverged from
a duplication.

Section 4.3.1:
Eq. (4.8) is really the simple pattern 1*3*...*(2m-5) in disguise.
The simple form also suggests how to derive it.

Algorithm 4.2. PGMA
The algorithm as stated does not specifically use arithmetic means.
In fact, it is a general hierarchical clustering algorithm.

Chapter 8:

Section 8.3, p 171:
- The g_ij are scalar products of the unknown coordinate vectors,
and the formula then follows from the familiar cosine rule.
- The Lagrange theorem is used to find the
distance from each point to the centroid, so it is needed if we want
to put the origin in the centroid (which is probably very reasonable).

Table 14.1, p. 287:
The numbers are numbers of different classes.


--
Last update 4 May
Eivind Coward