Why choose StrainSeeker for bacterial identification?
StrainSeeker's strengths
- Strain-level accuracy
- Very fast - 1 Gbp of data analyzed in minutes
- Comprehensive, easy-to-analyse results
- Can use custom, user-created databases
- Not based on single read analysis
- Can differentiate between known strains and their close relatives
StrainSeeker's limitations
- Custom database quality depends on guide tree
- Requires moderate to high coverage (0.1+) to identify strains
- Custom database building requires 200+ GB of HDD
Testing StrainSeeker on metagenomic samples
Sample 1 - large simulated metagenome consisting of more than 800 bacteria and archae
This sample was used in a study that compared 14 of the existing metagenome analysis tools View publicationStrainSeeker results | Samples (we used Set_A1 and Set_B1) | Community composition table
StrainSeeker results are averaged over the two sets. StrainSeeker's genus-level sensitivity was 0.968, precision 0.995. False negatives were mostly due to fact that these strains were not present in our database. Identification time was 12.3 minutes.
Sample 2 - mock metagenome consisting of 21 bacteria and a fungus
This sample was used in an early study that proposed to use k-mers for strain identification View publicationStrainSeeker results | Sample in FASTA format | Community composition table ('even' was used)
StrainSeeker accurately identified 20 bacterial strains (Actinomyces odontolyticus is not present in our database, nor are any fungi). Identification time was less than 5 minutes.