FAQs:
Q1: Do ALL the bacteria genera listed in browse page contain functional T3SSs?
A1: YES, but NOT ONLY. In current version of database, at least one strain from listed genus was validated to contain at least one functional T3SS.
Q2: Does each strain of the bacteria genera listed in browse page contain functional T3SSs?
A2: NO, but at least one strain from listed genus was validated to contain at least one functional T3SS. Some strains may not contain functional T3SSs.
Q3: Does each REPRESENTATIVE strain of the bacteria genera listed in browse page contain functional T3SSs?
A3: YES. For each representative strain, at least one functional T3SS was reported.
Q4: Rather than REPRESENTATIVE strains, are there ANY OTHER strains also containing functional T3SSs but not recorded?
A4: YES. There could be a number of strains OTHER THAN representative ones, which were not recorded in current version of database.
Q5: What are the criteria for inclusion of REPRESENTATIVE strains?
A5: The representative strains should be model strains, for which a number of T3SS research data were accumulated, and genome/T3SS_gene sequences were obtained.
Q6: For each representative strain, have ALL T3SS-related genes been annotated?
A6: NO. ONLY VALIDATED GENES were collected and annotated. There could be a lot of other T3SS-related genes not reported.
Q7: What does "T3 Othologs" mean?
A7: "T3 OTHOLOGS" represent a group of T3SS-related proteins with the same function and with the same evolutionary origin. T3 Othologs take genus and T3SS as unit, and therefore two genes belong to the same T3 ortholog cluster even if they are from different T3SSs of the same strain but with the same function and evolutionary origin.
Q8: Are the T3S effectors valdiated? Can I use them as T3S effector training dataset?
A8: YES, all the T3S effectors have been validated, and they can be used as training data. Before training, you may need to do homology filtering.
Q9: After I downloaded all the effector protein sequences, I found some proteins were annotated as 'APPARATUS' or 'TRANSLOCON'. What's wrong?
A9: It's NORMAL, since most translocon proteins are ALSO T3S secreted proteins and in our current version of database, there are only annotated as 'APPARATUS' or 'TRANSLOCON'.
Q10: How can I know the chaperone-effector pair relationship?
A10: You can download the table from DOWNLOAD page. In current version of database, the pair relationship between chaperones and effectors was also annotated in individual gene (effector/chaperone) annotation page.
Q11: For BPBAac, are there any specific requirements for input sequence format?
A11: YES. The input sequences must be FASTA format. The annotation line ('>xxxx') must be simple and with a single distinct identifcation without space (e.g., >seq1, ..., >seq100 ....).
Q12: For BPBAac, what do the different scores/thresholds represent?
A12: Different scores/threasholds represent different SELECTIVITY and SENSITIVITY. The score of 0.00 represents a selectivity of >95% and a sensitivity of >85% with original training dataset. 0.50 was selected as default threshold for a higher selectivity.
Q13: For T3SEpre, are there any specific requirements for input sequence format?
A13: YES. ALL the input sequences must be FASTA format. The annotation line ('>xxxx') must be simple and with a single distinct identifcation without space (e.g., >seq1, ..., >seq100 ....).
Q14: For T3SEpre, how many files or how many types of sequences should I imput?
A14: THREE: Amino Acid Sequence (Aac) file, Secondary Structure (Sse) File, and Solvent Accessibility (Acc) File. The three files should contain the SAME number of sequences. The sequence order and annotation ('>xxx') should be in accordance with each other in all 3 input files.
Q15: For T3SEpre, how should I prepare the Sse and Acc files?
A15: The Sse and Acc could be predicted with correspoding softwares with high accuracy (e.g., PSIPRED for Sse and SSpro for Acc). Sse should be represented with 3 elements ('C', 'H' and 'E' representing 'coil', 'helix' and 'strand' respectively) while Acc should be represented with 2 elements ('b' and 'e' representing 'buried' and 'exposed' respectively).
Q16: For T3SEpre, what do the different scores/thresholds represent?
A16: Different scores/threasholds represent different SELECTIVITY and SENSITIVITY. The score of 0.00 represents a selectivity of >97% and a sensitivity of >95% with original training dataset. 0.50 was selected as default threshold for a higher selectivity.
Q17: For T3_MM, what does the 'Probability' mean?
A17: If a protein is classified as a T3S sequence, the value describes the probability that the sequence to be T3S type; otherwise, when a protein is classified as a non-T3S sequence, the value describes the probability that the sequence to be non-T3S type.
Any question, please contact Yejun Wang: yejun.wang@gmail.com
Reference:
Wang Y, Huang H, Sun M, Zhang Q and Guo D. T3DB: an Integrated Database for Bacterial Type III Secretion System. BMC Bioinformatics. 2012, 13:66. Pubmed
Wang Y, Zhang Q, Sun M and Guo D. High-accuracy prediction of bacterial type III secreted effectors based on position-specific amino acid composition profiles. Bioinformatics. (2011) 27, 777-784. Pubmed
Wang Y, Sun M, Zhang Q and Guo D. Identification of New Type III Secreted Proteins Based on Position-specific Sequence-Structure Joint Features. Submitted.