pirovc / ganon_benchmark Goto Github PK

When running the "generate reference files" section, the output file from the top 3 command line is identical to the output file of the non-top 3 command. I'm using newly downloaded refseq sequences and a tab separated acc_len_taxid_assembly.txt file that consists of the RefSeq-Accn / Sequence-Length / Taxid / RefSeq Assembly Accession values from the updated_sequence_accession.txt file generated by genome_updater. It looks like the top 3 command line:

1. takes the RefSeq Unit Accession values from assembly.txt
2. sorts the values
3. removes duplicates
4. uses the values as input to zcat
5. appends the decompressed seqs to the t3a.fna file

Is the acc_len_taxid_assembly.txt not directly from the genome_updater aux file? Was there a previous filtering step I overlooked?

Sorry if I'm missing something obvious. So far it looks like ganon is a great solution to get around centrifuge's resource hungry index generation, thanks for the hard work!

Leo