HadrienG / 2019_classifiers_benchmark

Benchmarking of metagenomic classifiers
5 stars 0 forks source link

Run #9

Closed jhayer closed 4 years ago

jhayer commented 5 years ago

1. blast: do we run blastx as well? If yes, we need the prot DB 2. centrifuge: ok 3. diamond: ok 4. kaiju: ok 5. kraken: ok 6. kraken2: do we want to run the translated search? If yes we need prot DB 7. kslam: need to test to be sure about the tabulated output. Did not find option for num threads 8. mmseqs2: A. Is easy-search the command that we should use? B. We might need to run an LCA after. C. Do we run a translated search as well? If yes we need DB. 9. paladin: Does not support paired end, does not seem to have read type Illumina. Not sure about the options to use here 10. rapsearch: only outputs blast m8 format (text ASN.1) - we might need to perform LCA and get tabular output 11. salmon: quantification only. Output of quant.sf files. Do we want something else? 12. sourmash: I decided to use sourmash lca classify and sourmash lca summarize. Is that ok?