Hello,
I am processing eDNA metabarcoding data following this tutorial and customizing it to fit my needs.
I don’t understand what “maximum hits to consider/show” does in NCBI BLAST + blastn. Here’s why it’s confusing me:
Dataset #1 Round 1: maximum hits to consider/show → 1
filter sequence by id: I added my custom database and ncbi + blastn output
OUTPUT: 4 unique fish species.
Round 2: maximum hits to consider/show → 5
OUTPUT: 6 unique fish species.
Round 3: maximum hits to consider/show → 0 (no limit)
OUTPUT: 32 unique fish species (which seems too high for my local pond)
Dataset #2 Round 1: maximum hits to consider/show → 1
OUTPUT: 2 unique fish species.
Round 2: maximum hits to consider/show → 10
OUTPUT: 4 unique fish species.
Round 2: maximum hits to consider/show → 0
OUTPUT: 14 unique fish species.
I would like to set a firm, appropriate value for this parameter but I’m unsure what it truly represents.
Thank you for taking the time to read this, any insight would be greatly appreciated!
The option is limiting the search space during the alignment matching. The exact BLAST+ flags applied are slightly different based on the output option selected.
Maximum hits to consider/show *
Use zero for default limits. For HTML and plain text output this value is passed -num_descriptions and -num_alignments but for XML and tabular etc, this is used with -max_target_seqs instead. In either case, in addition to limiting the final output, this alters internal limits during the search, which can in some cases exclude matches which would otherwise become the best hit.
For your results, when you set this option to “1” are the reported hits “exact”? For metablast and tabular data, that usually means that the evalues are the same and are a “tie”.
For the other uses, the tool is likely reporting the strongest hit, then the weaker hits.
Some of the complications can come from the reference data (target database). The database in the tutorial was designed to reduce the chances of false cross-species or contamination hits.
This is complicated to explain but I hope this helps!
