Help

Below you can find a list of frequently asked questions as well as a list of help topics that are linked from the analysis and results pages. If you don't find all the help you need, please send an e-mail to one of the authors listed on the contact page.

FAQ

1. How can I generate input files for CAPS? → show

   The MEGA tool provides a very comfortable way to generate a collection of sequences from a range of species for your protein of interest and align them on your desktop. Alternatively you can use OMA to do this online.

2. How many sequences should be in the alignment / what evolutionary distance should be covered? → show
3. What to do if you get too many/few results? → show
   Lower the alpha threshold to decrease the number of results that get through the significance threshold. Increase the bootstrap value to show only those sites that change across a greater proportion of the phylogeny.
4. Why can't I see the Cytoscape output? → show

   You need Adobe Flash installed for the Cytoscape plug-in.

5. Why can't I see the 3D output? → show

   You need to have Java applets enabled in your web browser, check your browser for compatibility.
   On Apple computers add an exception for CAPS in the Java Control Panel (see here for more info.

Index

• Input format

  The user needs to upload a file with aligned DNA or protein sequences in Fasta, Mega or sequential Phylip format that represent homologues of the same protein/gene from different organisms. More information can be found here: File Formats

• Second data file

  For inter-molecular analysis a second data file can be submitted which contains homologues of a different gene/protein. The sequences in this file do not need to have the same number of species/strains but must have at least some common species/strains.

• PDB structure file

  For a more realistic and informative 3D graphic a PDB structure file can be provided. If the PDB ID is known then the server can retrieve these automatically.

• Tree files

  A Newick formatted tree can be submitted. We suggest using a tree calculated using either maximum likelihood or bayesian statistics using RAxML/MrBayes etc. In the case that a tree is not submitted CAPS2.0 will create a distance based tree automatically.

• Bootstrap threshold

  Increasing the bootstrap threshold leads to sites which are coevolving over a greater proportion of the phylogeny. Our advice is that if you wish to discover pairs of coevolving sites which are important for all species/strains in your alignment then this number should be close to 1. If you wish to identify lineage specific coevolution this number should be lower.

• Gap threshold

  This parameter allows the user to remove those columns of the alignment with large numbers of gaps. A value of 1.0 for this parameter will only include columns which have no gaps.

• Alpha value

  Type error I threshold to detect significant co-evolving sites. Lowering this value will result in fewer false positives but MAY remove true positives.

• Number of alignments to simulate

  This parameter defines how many multiple sequence alignments are simulated. We suggest 100 alignments. In the case that the user finds their data does not converge within 100 simulations we suggest using the convergence option (detailed below). Each simulated alignment will have the same number of columns as the input alignment and will be tested using the same method as the real data. This ensures that the data is compared to a null distribution without coevolution pressures.

• Time correction

  Using this option will scale BLOSUM scores evolutionary distance. These distances are either those supplied by the user in the phylogenetic tree OR are those JTT distances created when CAPS2 creates a phylogenetic tree for the user.

• Auto-retrieval of PDB

  CAPS searches through all sequences of structures in PDB for a match that has a minimum percentage identity to the reference sequence (90% by default). The best hit at or above the threshold is downloaded and used for the analysis.

• Convergence

  If after performing the simulations the distribution drawn has not converged within 0.0001 of the mean, additional similations will be performed until such time as there is convergence. We suggest using this when your data has many assymetries.

• Table view: Selection

  If a 3D structure was submitted a table will be included below the Jmol plugin, which allows to select individual groups of co-evolving sites via a radio button.

• Table view: Group name

  Each group of co-evolving sites is given a name consisting of incremental numbers.

• Table view: Mean distance

  For each pair of sites in a group that is found within the 3D structure the atomic distance is calculated. The mean of all pairs in a group is shown in the table.

• Table view: Sites

  This table column indicates the number of sites that the group of co-evolving sites consists of.

• Table view: 3D sites

  This table column indicates how many of the co-evolving sites are included in the 3D structure. This can be smaller than the overall number of sites if the 3D structure covers only parts of the input sequence.

• Table view: Residue indices

  This table column lists the residue indices associated with the co-evolving sites. These refer to the position of an aminoacid within the reference sequence. Bold numbers indicate residues that are found within the 3D structure.

• Table view: Residues

  This table column lists the residues associated with the co-evolving sites. Bold letters indicate residues that are found within the 3D structure.

For more help please feel free to contact the authors (see contact page).