note - FIZ Karlsruhe

The uploaded sequence text file becomes a query L-number
UPLOAD SUCCESSFULLY COMPLETED
L1 GENERATED
You may check if the sequence was uploaded correctly with D LQUE
=> D LQUE
Introduction to similarity searching
L1 ANSWER 1 DGENE COPYRIGHT 2012 THOMSON REUTERS on STN
LQUE tccagtgtgtccgctactccgctccccctcagtcctcagttcctcacctagcggtnnnnggcncgcggagacg
tagatggcggcttcggaggcggccaggcggcgcaacaccggtgacgagaggggacggtgatcgcgatccacag
cctggaggagtggagcatccagatcgaggaggccaacagcgccaagaagctggtggtgattgacttcactgca
acatggtgtcctccntnccgcnccatggctccaatttttnctgatatggccaagaagtccc
Running GETSIM and BLAST
Once the query sequence is uploaded the RUN GETSIM or RUN BLAST similarity search may be
started. To initiate the desired search the following field codes may be specified:
/SQP for searching peptide sequences [default]
/SQN for nucleotide sequences (Options: /SQN SIN, /SQN COM or /SQN BOTH)
/TSQN for searching peptide sequences against a nucleotide sequence database
For SQP and TSQN a peptide sequence is expected, for SQN a nucleotide sequence. If a mismatch
appears to be likely between the content of the query sequence and the field code, a warning is
issued. A GETSIM search is automatically adjusted to yield the most appropriate candidate
answers for the search type, query length and current size of the database. In contrast, the various
user-defined BLAST options provided by NCBI should be appended to the RUN BLAST
command as required. The syntax for including BLAST options is shown by example below.
Example
=> RUN BLAST L1 /SQP –F F –W 2 –E 10000 –M PAM30
The BLAST options which are most commonly adjusted by sequence searchers are as follows.
Low Complexity Filtering (on by default) (-F)
The low complexity filter can eliminate biologically uninteresting segments that have low
compositional complexity and are statistically significant, as determined by specific programs for
peptide or nucleotide sequences in nature. Filtering is applied to the query sequence and is
indicated by a series of Xs for peptide sequences and Ns for nucleotide sequences. Low
complexity filtering can be turned off (i.e. set to F for False).
Word Size (-W)
Word Size is the length of the character string fragments of a sequence query that are used as the
basis for a BLAST search. For SQN the default is 11 and the range 7-23. For all other BLAST
searches the default is 3 and the range 2-3. For short search queries, reducing the default word
size can give improved search results.
GENESEQ on STN (DGENE) Workshop Manual | Page 15