Case 1: From a UniProt ID to KEGG information

This walkthrough starts from a protein identifier in an outside database (UniProt / Swiss-Prot) and pulls together the corresponding KEGG gene, its sequences, orthology group, reactions and pathways.

1. Convert an outside identifier to a KEGG identifier

kegg_conv maps identifiers between KEGG and outside databases. Some common gene / protein identifier prefixes are:

Only identifiers with a hit in KEGG are returned:

conv = KEGGAPI.kegg_conv("genes", "uniprot:A0A072UR65")
DataFrame(conv.data, conv.colnames)

Several identifiers from the same database can be converted in one call by passing a vector (for example a column read from a CSV file with CSV.jl). The input is automatically split into chunks of 10 to respect the KEGG rate limit:

using CSV

df = DataFrame(CSV.File("subset_data.csv"))
entries = string.("uniprot:", df.Entry)
conv = KEGGAPI.kegg_conv("genes", entries)
DataFrame(conv.data, conv.colnames)

The reverse direction (KEGG → outside database) works the same way:

conv = KEGGAPI.kegg_conv("ncbi-proteinid", "mtr:25493984")
DataFrame(conv.data, conv.colnames)

2. Retrieve the gene entry

Use kegg_get to fetch the full flat-file entry for a KEGG gene. A single entry is returned as a String in the .data field:

gene = KEGGAPI.kegg_get("mtr:25493984")
println(join(first(split(gene.data, "\n"), 8), "\n"))

ENTRY       25493984          CDS       T01716
NAME        (RefSeq) class V chitinase CHIT5b
ORTHOLOGY   K01183  chitinase [EC:3.2.1.14]
ORGANISM    mtr  Medicago truncatula (barrel medic)
PATHWAY     mtr00520  Amino sugar and nucleotide sugar metabolism
            mtr01100  Metabolic pathways
BRITE       KEGG Orthology (KO) [BR:mtr00001]
             09100 Metabolism

3. Download sequences

kegg_get can return amino-acid (:aaseq) or nucleotide (:ntseq) FASTA sequences. .data holds one FASTA record per requested entry, which you can write to a file (e.g. with FastaIO.jl):

using FastaIO

seqs = KEGGAPI.kegg_get(["mtr:25493984", "shz:shn_30305"], :aaseq)
FastaWriter("aaseq.fasta") do fw
    for record in seqs.data
        write(fw, record)
    end
end

4. Orthology, reactions and pathways

kegg_link finds cross-references between databases. The orthology (KO) group for the gene:

ko = KEGGAPI.kegg_link("ko", "mtr:25493984")
DataFrame(ko.data, ko.colnames)

Reactions associated with that ortholog:

rxns = KEGGAPI.kegg_link("reaction", "K01183")
DataFrame(rxns.data, rxns.colnames)

Pathways the gene participates in:

paths = KEGGAPI.kegg_link("pathway", "mtr:25493984")
DataFrame(paths.data, paths.colnames)

5. All genes in an orthology group

The same kegg_link call, targeting genes, expands an orthology group into every member gene across organisms:

ko_genes = KEGGAPI.kegg_link("genes", "K01183")
first(DataFrame(ko_genes.data, ko_genes.colnames), 5)

The second column of ko_genes.data is a vector of gene identifiers that can be fed straight back into kegg_get(...; :aaseq) or :ntseq to build, for example, a multiple-sequence-alignment input.

6. Download a pathway map

The :image option returns the PNG bytes of a pathway map, which you can write to disk and open with your favourite image viewer or Images.jl:

img = KEGGAPI.kegg_get("map00520", :image)
open("map00520.png", "w") do io
    write(io, img.data)
end