NrRecipes

UserPreferences

Recipes related to the Non-redundant (NR) database of sequences

1. Rebuilding the NR, e.g., after adding new organisms.

NOTE: There was a chage in the format of the NR at some point that introduced id's of the form "xxx*" The code assumes this format, so if you have an older NR and peg.syn file, you must change them to this new format first. See recipe #2]

• mkdir workdir
• cd workdir
• cp -p ~fig/FIGdisk/FIGData/Global/nr .
• cp -p ~fig/FIGdisk/FIGData/Global/peg.synonums .
• build_nr_sources > nr_sources
• build_nr nr_sources ./nr ./peg.synonyms newNr newPeg.synonyms
• [copy newNr and newPeg.synonyms back to Global if you are happy with them.]
• load_peg_mapping

Note- nr and peg.synonyms are present as inputs to preserve id's from run to run and to determine the next free ID. If you are *only* doing build_nr to compare to genomes and have no intention of ever installing the results, then you can get /dev/null for these inputs and speed up the processing.

2. Converting old NR to new NR format

• assign_synonym_ids old.peg.syn new.peg.syn id.map
• assign_synonym_ids_to_nr old.nr id.map new.nr
• assign_syn_ids_to_sims

3. Finding the list of sequences that are new (newNr vs. oldNR):

• compute_changed_ids_for_nrs oldnr oldpegsyn newnr added-out changed-out deleted-out

4. Getting fasta info for all the new sequences (see #3)

• pull_fasta_entries new-nr < added-out