-i filename: Enter the path to the file containing gene family data. The data file format must be tab-delimited with UNIX line endings. Family description may contain spaces (but not tabs). The first line must contain labels in the order: Description ID, A list of tab-delimited taxon names
The taxon names must be spelled exactly as they are in the provided phylogenetic tree. Each subsequent line then corresponds to a single gene family. If the data file contains taxa that do not appear in the tree structure, they are not considered in the analysis.
Here is an example of an input data file:
If the file is loaded correctly, CAFE will output summary information about the current data file to the log file.
-t # of CPU threads: The maximum number (integer) of CPU threads to be used. Default: 8.
-l filename: Enter the path to the file where CAFE will write the main output. This file will contain a summary of input parameters as well as details of λ searches, including likelihood scores and maximum likelihood values of λ. If the file does not exist, CAFE will create it for you; if the file already exists, CAFE will append the results to the previous file. Default: output to screen (no log file created).
-p α: For each family in the data file, CAFE computes a probability (p-value) of observing the data given the average rate of gain and loss of genes. All else being equal, families with more variance in size are expected to have lower p-values. The significance level (α, a float) allows the user to specify the cutoff for subsequent analyses. Families with p-values larger than the designated significance level will not be included in the identification of the most unlikely branch. Default: 0.01.
-r # of random samples: To determine the probability of a gene family with the observed sizes among taxa, CAFE uses a Monte Carlo re-sampling procedure. This option specifies the number of samples CAFE should use to calculate p-values. The tradeoff is between precision and computation time; in most cases 1000 samples should provide reasonable balance. Default: 1000.
-filter: The birth-death model of CAFE assumes at least one gene in the root of the species tree. This assumption may not be valid for families that were created after the most recent common ancestor of all species. The filter option filters out the families that are inferred (by parsimony) to have no genes in the root node of the species tree.