I love Dan Gusfield’s book, “Algorithms on Strings, Trees, and Sequences”. While it’s also a bit out of date, it covers roughly complementary topics to Durbin et al., including all the basic sequence alignment algorithms, including suffix trees, which form the logical basis of Burrows-Wheeler alignment. But you’ll need to understand something like Cormen et al.’s algorithms book as a prerequisite.

I also loved Alberts et al. “Molecular Biology of the Cell” book. I’m a stats algorithms person who mainly works in computational linguistics, but I found it easy to read. I read this with only high school chemistry as a background. You really have to read the first section of that, or something similar, to understand the problems. And it’s a joy to read!

Nothing out there really helps with the kind of factor models that are interesting for statistical modeling for, say, expression, such as was used in d-Chip for microarrays or some of the RNA-Seq stuff coming out now. My fave book in that area is Gelman and Hill’s “Data Analysis Using Regression and Multilevel/Hierarchical Models”, or at a more advanced level, Gelman et al.’s “Bayesian Data Analysis”, but they have no sequence data, and only the occassional biology example. The Bishop or Hastie et al. machine learning books also cover some of this same material, with the advantage of covering all the non-probabilistic approaches like SVMs. These all presuppose you have basic math stats down, such as found in Degroot and Schervish or in Larson and Marx.