Downloads reference sequences from NCBI for given organism(s) as specified by the input taxon id(s). The records are mapped to gene identifiers and reduced to this level. If there are multiple protein accessions that map to the same gene identifier, then the highest NP accession number is used if available, otherwise the highest XP accession number is used. This mapping is performed through NCBI’s gene2accession file. Additionally, homologene information is included if available.
Here is an example of what a header looks like in the output.
>geneid|91703|ref|NP_542389.1|taxon|9606|gi|18087825|homologene|12401| N-acyl-aromatic-L-amino acid amidohydrolase (carboxylate-forming) [Homo sapiens]
python3, pandas, numpy
git clone https://github.com/asalt/RefProtDB.git
cd RefProtDB
pip install .Running the program is straightforward:
RefProtDB --outdir /path/to/save/output/ 9606 10090 > programlog.txt # this last part is optional3 files will be generated here.
A combined fasta file with both taxa, a fasta file with only human, and a
fasta file with only mouse.
If this splitting is not desired, the --no-split flag can be passed.