Eff3ctidor

Select an input ORFs file:

Select an input effectors file (optional):

Include additional positive samples based on sequence similarity to the T3E dataset?
Yes No

E-mail:

Job title:

Would you like to include a prediction of the secretion signal in your analysis?
Yes No

Minimal identity percentage of orthologs and paralogs:

Minimal coverage percentage of orthologs and paralogs:

Upload effectors file for homology search:

Upload a host proteome archive:

Upload an archive with proteome records of closely related bacteria without T3SS:

Upload a GFF3 file:

To search for regulatory motifs in the promoters of the genes please supply the following data:

Upload a full genome FASTA file:

Select regulatory elements to be searched for in the promoters:

PIP-box
hrp-box
mxiE-box
exs-box
tts-box

Input specification

The input for Effectidor is a text file with all the Open Reading Frames (ORFs) sequences of the analyzed genome(s) - both from the chromosome and plasmid(s) when available, even if per your knowledge the T3SS and T3Es are found only on one of them. The file should be in a DNA FASTA format.
You can find instructions to download this file here.
If you wish to analyzed multiple genomes, the ORFs from each genome should be supplied in separate files and all the files should be compressed together in a ZIP archive.
Make sure the names of the fasta files are indicative for you, as these are the names that will appear in the results. For example, 'genome1.fasta','genome2.fasta'.
Please make sure all the files packed in your input zip archive are in DNA FASTA format, and are not packed in directories - Directories within the zip archive are ignored.

Note, only protein coding genes should be included in this input. Non-coding RNA genes, such as tRNA, should not be included. If you follow the instructions above, no such genes are expected in your input file. Alternatively, if such RNA genes do appear in your data and you supply a GFF3 file indicating it (in the advanced options), we will filter them out automatically.

Input specification

This file should be in a DNA FASTA format. The records supplied in this file must be copied from the main ORFs file, and will be treated as "known effectors" to train the machine-learning algorithm.
When analyzing multiple genomes, you can create such effectors file for each (or some) of the analyzed genomes. In this case, please pack these files in a ZIP archive. Make sure the names of these files are matching to the corresponding ORFs file. For example, if the analyzed ORFs files are 'genome1.fasta','genome2.fasta', the names of the corresponding effectors files should be 'genome1.fna','genome2.fna' (the file extension does not matter).
This is an optional (and recommended) input. If you choose not to supply it for a certain genome, the records that show high sequence similarity to our effector dataset (see "Data") will be used as the "known effectors" in this genome. If one or more of these "known effectors" are clustered in the same OG with other genes in the genomes for which you supplied this input, the entire OG will be used as "known effector" for the training.
Please note that the predictions are conducted after training the machine-learning algorithm on these "known effectors" of the given genomes, and no exterior effectors' data are used for training. Thus, it is most recommended adding these records to achieve better predictions of potential novel T3Es. This is especially important when not many effectors are known in the bacterial genome you are analyzing.

Input specification

The previous effectors input (from the analyzed genome(s)) is optional, and if not supplied we build the effectors set based on sequence similarity to an established effector dataset. Since you supplied an effectors input, we can either use it as the only positive set, or we can add to it ORFs from your input that share high sequence similarity to our effector dataset. Choose "yes" if you want us to search for additional effectors based on sequence similarity to use for the training, and "no" if you want your input to be the only data used in the positive set.

Please note that sequence similarity to our effector data is used as a feature in any case.

Input specification

The minimal identity percentage required for a pair of best reciprocal hits to be considered as orthologs (lower values will be filtered out).

Input specification

The minimal coverage percentage required for a pair of best reciprocal hits to be considered as orthologs (lower values will be filtered out).

Input specification

In this input you may upload a protein FASTA file with type 3 effector records of other bacteria to be used in addition to our effector dataset for homology searches.

Input specification

In this input you may upload host proteome(s).
You should upload protein FASTA file(s), compressed together in a zip archive.
The files should not be packed inside directories - directories packed in the zip archive will be ignored. Also files with names starting with a dot (.) or an underscore (_) will be ignored.

If you upload this archive, we will search for sequence similarities between the protein sequences in the compressed FASTA files and the ORFs in your input genome(s).
Sequence similarity between bacterial proteins and host proteins is potentially a very important feature for the classification as it captures Eukaryotic domains within the bacterial proteins.

Input specification

In this input you may upload proteome(s) of closely related bacteria that do not encode a T3SS.
You should upload protein FASTA files, such that each FASTA file contains all the protein sequences of a single bacterium, compressed together in a zip archive.
The files should not be packed inside directories - directories packed in the zip archive will be ignored. Also files with names starting with a dot (.) or an underscore (_) will be ignored.

These data will be used for sequence similarity search. Each FASTA file within the zip archive, representing a different proteome, will be used for a different search. Sequence similarity to proteins of closely related bacteria without a T3SS is a strong negative feature for T3Es classification.

Input specification

GFF3 files will be used for extraction of genome organization features.
If you wish to analyze multiple genomes, such a file should be given for each genome with matching names, for example if the ORFs files are named 'genome1.fasta' and 'genome2.fasta', the corresponding names of the GFF files should be 'genome1.gff3' and 'genome2.gff3' (note that the basename should match, the file extension does not matter). In this case, these files should be compressed in a single zip archive.
The files within the zip archive should not be inside directories - directories packed in the zip archive will be ignored. Also files with names starting with a dot (.) or an underscore (_), or with names not found in the ORFs files will be ignored.

Input specification

These files will be used to search for cis-regulatory elements in the promoters of the genes, according to your selection.
The regulatory elements available for search are PIP-box, hrp-box, mxiE-box, exs-box, and tts-box.

If you wish to upload more than one file (to analyze multiple genomes), compress them in a single zip archive.
In this case, make sure the names of the files are matching to the ORFs and gff files. For example, if the ORFs files are named 'genome1.fasta' and genome2.fasta', the names of the corresponding gff files are 'genome1.gff3' and genome2.gff3', then the corresponding full genome files should be named 'genome1.txt' and 'genome2.txt' (note that the file extension is just an example, and is of no importance).
The files should not be packed inside directories - directories packed in the zip archive will be ignored. Also files with names starting with a dot (.) or an underscore (_), or with names not found in the ORFs files will be ignored.

Input specification

PIP-box is relevant for Xanthomonas, Ralstonia, and Acidovorax.
hrp-box is relevant for Pseudomonas syringae and plant pathogens of the family of Enterobacteria.
mxiE-box is relevant for Shigella.
exs-box is relevant for Pseudomonas aeruginosa.
tts-box is relevant for rhizobia.

Make sure to choose only regulatory elements that are relevant for the bacterium you analyze. Non-relevant regulatory elements may hamper the classification accuracy.

Input specification

This feature can help us find potential T3Es. The disadvantage of this inclusion: longer running time of Effectidor.
More information is available in this paper. Please refer to this paper if you use this feature in your analysis.

Input specification

The disadvantage of this inclusion: longer running time of Effectidor.
More information about this prediction is available in this paper.

Research Site | Pupko Group