Two separate but related projects make up the work of this thesis. The growing amount of sequence data available in public databases provides an opportunity to compare species in new ways. It can be shown that there is a systematic change in amino acid composition in a dataset of sequences from 69 species possessing a range of optimal growth temperatures. By creating a phylogenetic tree of all available Archaea, pairs may be selected that contain a relatively closely related mesophile and (hyper)thermophile. In addition, pairs may be selected from Bacteria to include psychrophiles as well as other thermophiles. An evolutionary model is derived here that detects amino acid asymmetries in these species pairs beyond what might be expected to be caused by differences in GC content. This amino acid asymmetry can then be plausibly explained by temperature adaptation occurring in these species since they diverged from a common ancestor. In the second part, similarity searches using molecular sequences are drawn as networks, where open reading frames in one species may be linked to a corresponding sequence in another species if the similarity search score is above a given threshold. This process is similar to that used to identify orthologous sequences for use in evolutionary models. When drawn as a network of distinct clusters of similarity, patterns emerge that can be spurious or have some biological relevance. This work identifies the need to develop better methods of analyzing these network clusters. / Thesis / Master of Science (MSc)