How to download CDS from ncbi for pangenome analysis?

Question

[sidraas]

How to download *_cds_from_genomic.fna.gz (CDS from genomic FASTA) from ncbi for pangenome analysis?

[Ram]

What have you tried? Please add more detail to your question.

Answer 1

[Carlos Caicedo]

You can use ncbi-genome-download. The instructions for installation and usage are available here

https://github.com/kblin/ncbi-genome-download

For what you are asking, you need to specify the -F parameter as 'cds-fasta'

Answer 2

[MirianT_NCBI]

Hi,

You can use NCBI Datasets. The default genome package includes:

• genomic fasta (chr*.fna, unplaced.scaf.fna)
  • transcript fasta (rna.fna)
  • protein fasta (protein.faa)
  • CDS fasta (cds_from_genomic.fna)
  • GFF3 (genomic.gff)
  • metadata files (sequence_report.jsonl for each assembly, assembly_data_report.jsonl and dataset_catalog.json).

To download only the cds, you can use the following command (I'm using human as example, but you can use any taxonomic level):

datasets download genome taxon human \
--exclude-gff3 --exclude-protein --exclude-rna --exclude-seq \
--filename cds_only.zip

If you're downloading a really large number of files (let's say all vertebrates), I would recommend adding the flag --dehydrated. With this flag, datasets downloads the json and jsonl files, and a file called fetch.txt with paths to the data to be downloaded (rehydrated). To rehydrate a package, you can follow the steps below:

unzip cds_only.zip -d cds_only
datasets rehydrate --directory cds_only

I hope it helps!