Hot-spring chemistry and model proteins from a metagenome

This vignette for the CHNOSZ package demonstrates calculations of the relative stabilities of proteins with amino acid compositions derived from metagenomic data in a hot spring. The calculations are described in more detail in a recent paper [1]. The purpose of this document is to present a minimal amount of code to show how most of the figures in the paper can be generated. The code actually used to make the figures in the paper is available online as supporting information for the paper. “Bison Pool” is a hot spring in Sentinel Meadows, in the Lower Geyser Basin of Yellowstone National Park. Environmental DNA sequences from biofilm samples at five sites along the outflow are available from the Joint Genome Institute’s IMG/M system [2]. IMG/M also provides the sequences of proteins inferred from the metagenome, with predicted functional annotations. The sample IDs used in the field, and also in IMG/M, have the letters N, S, R, Q, P, which refer to sites 1, 2, 3, 4, 5, in that order; site 1 is the boiling source of the hot spring and site 5 is ∼22 meters downstream where the water on some days was in the range of ∼55-60 ◦C [3]. The amino acid compositions of the model proteins are derived from the metagenome. The “overall” model proteins have amino acid compositions that are the average of all protein sequences at any sampling site. There are also 20 different “classified” model proteins for subsets of the metagenome that share keywords in the annotations.