Protein Pattern Recognition Services
BlastP : LINK
- blastp with Non-redundant protein sequences(nr): ABL1_HUMAN (left bottom figure)
CDD (LINK) and CD-Search (link)
- Take a look on the prediction results of conserved domains. Take cd11850 as an example to use conserved domain databases. Following is a partial result.
There is plenty of information noting this domain, such as phylogeny tree (the above and middle figure) built with the components in the databases, sequence alignment result (the above and right figure), etc.
Try CD-search in FASTA format sequence data. Take ABL1_HUMAN as an example.
- The following is searching result from CD-search operation under the input ABL1_HUMAN. The green label represents the PTKc_Ab1 surrounded by red rectangle. The result is equal to the result from BLASTP operations.
ScanProsite : LINK
- Scan patterns through ScanProsite and take ABL_HUMAN as an example. The following is a partial image showing the input and the parameters.
- The parameters include the scanning mode, high probability or high sensitivity, and the output format, graphical view, text, table or FASTA, etc. The following is the partial result from the scanning pattern under the input ABL1_HUMAN. The results consist of the geographic location of each pattern on the sequence and the detail description of each pattern (not shown).
Pfam : LINK
- Search the family through Pfam and take the protein ABL1_HUMAN as an example.
MAGIIC-PRO : LINK
- Functional region prediction through MAGIIC-PRO.
- The left and upper figure is the input to undergo the prediction of structure and function. The right and upper figure is the step to show several sequence similar with the input sequence. And these sequences would be as training data in order to further predict and to modify several parameters in operation processes. Change the training data if it is necessary. The input sequence with many training sequences might go on a prediction of several aspects, such as functional regions and structural patterns. The partial results are the bottom figures. It showed lots of information, such as the conserved amino acids in the patterns. (Refer to the right and lower figure)
Compare
- The compares among several servers mentioned above:
| Items & Functions | BLAST (-P) | CDD | ScanProsite | Pfam | MAGIIC-Pro |
|---|---|---|---|---|---|
| Source | NCBI (USA) | NCBI (USA) | ExPASy (SIB) | Trust Sanger Institute (UK) | BIME, NTU (Taiwan) |
| Multiple parameters | Yes (many) | Yes (less)$$^{[1]}$$ | Yes (less)$$^{[2]}$$ | No$$^{[3]}$$ | Yes(many)$$^{[4]}$$ |
| Sequence alignments & statistical data | Yes$$^{[5]}$$ | Yes | No | No | No |
| Pattern information | Yes | Yes | Yes$$^{[6]}$$ | Yes | Yes$$^{[7]}$$ |
| Phylogeny tree | No | Yes | No | No | Partial$$^{[8]}$$ |
| 3D-Structure information | No | No | No | No | Yes$$^{[9]}$$ |
- Only basic operations: databases, expect value, etc.
- Scanning modes, probability or sensitivity, and output format, graphic view or FASTA, etc.
- Only the name for searching: not a good design for only specific name searching
- Include many detailed settings for multiple usages, such as identity cut off and the matrix.
- The statistical result includes e-value, identity, scores, etc.
- The special is that ScanProsite provides information of sites, such as the bisulfide bridge, active site, etc.
- The special is calculating and marking the conserved amino acids in the patterns among several sequences of different species.
- Based on the result the relation among several species could be inferred, but not likely CDD showing the total phylogeny tree.
- Another special aspect is that the MAGIIC-Pro server could show 3D structure and provide for operations.