Table 1 Tentacle applied to three use cases. All use cases used the same read data consisting of 1,238,598,682 reads with a total size of 407 GiB in compressed FASTQ (2,213 GiB uncompressed) [6]. The examples were run on 30 nodes, with the cluster system login node hosting the master process. The following options were used, pBLAT: -threads=16 -minIdentity=90 -out=blast8; GEM: -T 16 -m 0.04 -e 0.04 –min-matched-bases 0.80 –granularity 2500000; USEARCH: -usearch_local -query_cov 1.0 -id 0.9 -blast6out
From: Tentacle: distributed quantification of genes in metagenomes
Use case | 1 | 2 | 3 |
|---|---|---|---|
Reads mapped to their contigs | Reads mapped to large DB | Reads mapped to peptide DB | |
Mapper | pBLAT | GEM | USEARCH |
Type of reference | Per sample contigs (nucleotide) [6] | BGI Refseq geneset (nucleotide) [6] | Resqu; antibiotic resistance gene |
database (peptide) [53] | |||
Reference size (bytes) | approx. 160 MiB per sample | 3.0 GiB | 1 MiB |
Reference size (sequences) | 6,589,348 | 3,305,138 | 3,019 |
Runtime (core hours) | 720 | 3,072 | 296 |
Runtime (wall-clock) | 1h 30m | 6h 24m | 0h 37m |