AMPK Methods and Protocols

particular tool over the one that we have listed, it is easy to substi-

tute it in the analysis workflow (Fig.1). All analyses can be run on a

typical desktop computer with at least 8 GB RAM and 500 GB disk

space. We suggest any Linux distribution or alternatively Mac OSX

as operating systems. Make sure that Perl (v5.22.1), Python

(v2.7), JAVA (v1.7), and R (v3.3.3) are installed.

2.1 Protein Sequence
Databases

You can generate a data collection optimized for your analysis from

a plethora of different public web pages and data repositories. Note

that almost all data is generated from draft genome sequences.

Thus, a fraction of the genome is not represented in the assembly,

and the annotation of genes is likely to be incomplete. Therefore, a

fraction of genes that are encoded in a species’ genome might not

be represented in the reconstructed gene set.

2.1.1 Joint Genome
Institute (JGI) Genome
Portal [14]

The JGI genome portal (http://genome.jgi.doe.gov) provides

access to all JGI genome databases. The complete gene sets for

individual species from all areas of the tree of life are available for

download through an ftp server (seeNote 1). The JGIGenome

OnLineDatabase (https://gold.jgi.doe.gov/) provides an excel-

lent start for compiling a data set.

2.1.2 National Center
for Biotechnology
Information (NCBI)

NCBI (https://www.ncbi.nlm.nih.gov/) provides an open-source

web server for searching and retrieving protein and nucleotide

KEGG Pathway

Proteins

OMA

OrthoDB

HaMStR-OneSeq

FAS-S

PhyloProfile

Proteins

Species

Tree Reconstruction

(RAxML)

Phylogenetic

Hypothesis Testing

(SH Test)

A

B

B'

C

D

E

F

Fig. 1Example workflow for tracing the evolutionary history of a metabolic pathway. (a) Identify the pathway
of interest and retrieve the protein sequences for the individual pathways components from the KEGG
database. (b) For every protein, retrieve orthologs in a selection of target taxa from public databases. (b^0 )
Alternatively identify orthologs using a targeted ortholog search with HaMStR-OneSeq. (c) Compute for each
ortholog its pairwise feature architecture similarity (FAS) to the seed protein. Note that this step is already
integrated into the HaMStR-OneSeq search. (d) Create a phylogenetic profile displaying orthologs and FAS
information for all pathway components with PhyloProfile. (e) Reconstruct the evolutionary relationships for
individual or collections of seed proteins with RAxML. (f) Optionally, test the resulting sequence tree against an
expected phylogeny, e.g., the species tree

Tracing AMPK Evolution 113