Host prediction

[snayfach]

Hadza identifiers don't match master table: Nepal_MoBio_Fiber-Hadza-Nepal_B_5_THA1056YZ.71 CRISPR spacers to all 3 databases kmers to only 2/3 databases

MySQL tables

• host_taxonomy: only has GTDB r202 (258406) / UHGG v2.0 (289232)
• hadza_mags: 54779 and includes mapping to metadata table
• host_genomes: this is the table that Bryan provided me for the genomes he's using
  • uses GTDB r207**
• otu_to_host_spacers: contains matches to all 3 databases
• otu_to_host_kmers: not updated with Hadza matches

To Do

• [x] rerun host prediction using correct input data (Hadza, GTDB-gut, UHGG)
• [x] export host prediction tables
• [x] export crispr spacers
• [x] export list of hosts and their taxonomy

Genome info counts (mysql = host_genomes):

• GTDB (123,500)
• UHGG (286,387 vs 289,232)
• Hadza (54,779)

host_taxonomy.tsv file: looks good!

• GTDB (123,500**)
• UHGG (286,387**)
• Hadza (?? prob right number)

Do I have hits to UHGG genomes absent from list?

[snayfach]

Reuse text from IMG/VR: https://academic.oup.com/nar/advance-article/doi/10.1093/nar/gkac1037/6833254