Prot.accession2txid导入问题

Anupam · November 24, 2024, 6:57pm

please have a look at these threads, and than you can decide

Optimal configuration for metagenome assembly classification

opened 03:39PM - 21 Sep 18 UTC

closed 01:21PM - 22 Sep 18 UTC

I'm using Diamond for classifying metagenome contigs (HiSeq/MetaSpades) against …NR using a high memory server, and it works great! This workflow is so much easier than it used to be thanks to recent updates. However I suspect the sensitivity of some of the classifications for viral contigs are compromised by hits against multiple protein subjects. This is causing many long contigs to only be labelled at family level. I have reviewed the documentation extensively but remain unclear on whether I should use `--range-culling`, or whether this is even possible given that frameshift alignment mode seems inappropriate here, with me using assembled contigs (from accurate short reads) as queries. Perhaps you could recommend settings for sensitive taxonomic classification of viral and bacterial contigs? Current usage: ``` ~/diamond/0.9.22/diamond blastx \ --query <> \ --db <> \ --out <> \ --more-sensitive \ --outfmt 102 \ --block-size 20 \ --index-chunks 1 \ --tmpdir /dev/shm \ --threads 16 ```