About MicrobesOnline Sequence Search

To search for a sequence in MicrobesOnline, click here.

Protein search

1. MicrobesOnline sequence search first looks for near-exact matches by using BLAT, which is very fast. This can find sequences that match protein queries even if no gene was annotated.
2. Next, if you request it, sequence search uses FastHMM to find domains & families. (It searches for hits to COG, Pfam, TIGRFam, Superfamily, Gene3D, and PIRSF, but not Panther.)
3. Finally, sequence search selects potential homologs and uses BLASTp to compare them to the query. Whenever possible, it uses domains and families to quickly find potential homologs; otherwise it uses BLAST against sequences in our database that do not belong to families. Given these potential homologs, FastBLAST uses BLASTp to compare them to the query and to rank the hits.

Sequences should be more than 25 amino acids in length for best results. MicrobesOnline will search the database for shorter queries, but will not filter hits on e-value in order to show all hits. Queries shorter than approximately 15 amino acids may not produce good (or any) results; you should try to use a longer query if possible. You can also BLAST against MicrobesOnline databases, but BLAST is much slower than sequence search.

Nucleotide search

The nucleotide search uses MEGABLAST against all of the genomes in the MicrobesOnline database. Queries shorter than 30 nucleotides will use BLASTN instead. Queries shorter than 20 nucleotides may not produce good (or any) results; you should try to use a longer query if possible. You can also BLAST against MicrobesOnline databases instead if this is not sensitive enough.