Database Selection
SEARCH MODE:
Full DB Mode
This is an exhaustive search, and more likely to find very remote homologues than Motiftracker, but it can take as long as ten minutes to return a result.
Motiftracker Mode
This is a fast search (average one minute), as the clustered database set is first pruned to exclude clusters where query-residue frequency is under 50%.
DATABASE:
pdb
Search against the PDB - a database of experimentally solved protein structures. The date of the last PDB snapshot is given at the top of the start page. Note that the PDB target sequence will not contain any unmodelled residues.
scop
Search against the scop40 dataset of the current SCOP release. The SCOP version used is given at the top of the start page. Please be aware that some SCOP domains are non-contiguous, and consist of conjoined segments of the underlying PDB sequence.
nrdb40
Search against nrdb40, a representative set of all known proteins at 40% redundancy from our in-house, non-redundant sequence database, PairsDB. The ADDA Server is based on PairsDB, and you can use it to browse your hit proteins and view family information. Version information given at top of start page.
pfama
Search against the Pfam-A dataset, mapped onto nrdb40. Version information given at top of start page.
Query Protein
Choose your input in one of three ways - by identifier, nid or sequence:
Identifier
Enter a Uniprot, TrEMBL, PIR, NCBI or other identifier or accession number into the first input box. If we can locate it in our database, GTG will run, otherwise a warning message will be returned.
nid
Our non-redundant sequence database, PairsDB, uses unique, stable numerical identifiers (nids) for protein sequences. If you know your protein nid, which you can find by searching ADDA or PairsDB (coming soon), enter it in the second input box. You may also specify a sequence range for your nid using underscore-separated values, eg: 123456_1_120.
Sequence
When you submit a sequence, a fast Fragment search is run first to look for an exact match in our database. If none is found, then a Blast search is performed, which can take up to 30 seconds to return a result. Please do not re-click submit while the search is running. The sequence should be in plain or fasta format. Numbers and whitespace will be stripped out.
Parameters
There are a few parameters you can vary which affect the search. The default values are a good set to use if you're unsure about changing them. Here is a little more information on each:
Full Hits Returned
The number of homologues you want to search for if running Full DB mode, up to a maximum of 50. Note that the higher this number, the more time it will take to return a result.
Motiftracker Hits Returned
The number of homologues you want to search for if running Motiftracker mode, up to a maximum of 100. Note that the higher this number, the more time it will take to return a result.
Gap Penalties
Whether or not to penalize gaps in the alignment. Note that these are not equivalent to Needle/Water gap penalties; the effect is more cosmetic. If your resulting alignments are fragmented or have long gaps near the ends, setting penalties can help to tidy them up. Set this to «off» to leave all alignments unaltered.
Score Weighting
The GTG uses two scores: Consistency, a score based on neighbourhood overlap along a trace; and Motif, a weighted identity score. This parameter determines what proportion of each is used to obtain the Overall score, eg, 20:80 means Overall score equals 20% Consistency + 80% Motif scores.