## DormanResearch : CBrotherManual

Referers: cBrother :: (Remote :: Orphans :: Tree )
Dorman Wiki
Dorman Lab Wiki

### 0cBrother Manual

1. Preliminaries
2. Tutorial or Worked Examples
3. Data
4. Priors
1. Topology
1. Uniform: default, uninformative prior on topologies
2. Monophyly: test for independent recombination events for recombinants with same structure
3. Event: test for independent recombination events at cops shared by multiple recombinants
2. Change points
5. Processing cBrother Output
6. Command File Options: The command file options are listed more-or-less in the order of greatest utility.  If links are not included, the description here should suffice
1. Controlling how cBrother runs: Use these commands to set the properties of the MCMC run.
1. length: set the length of the MCMC run
2. burnin: set the length of the MCMC burnin
3. subsample: set the subsample rate for the MCMC run
4. continue: set to continue a previous MCMC run (part of checkpointing)
5. cmdfile_errors: determine how errors in the command file format will be reported
2. Controlling the model and topology space cBrother uses
1. model: select the type of model/analysis
2. change_points: set the types of allowable change points
3. allow_split_ancestry: allow or exclude certain types of topologies from the list of possibilities
4. allow_ancient_ancestry: allow or exclude certain types of topologies from the list of possibilities
5. parent_tree: fixed genotype tree
6. number_of_clades: specify the number of genotypes in the data
7. clade: define the sequences belonging to a genotype
3. Controlling what cBrother outputs
1. output_style: set the output style
2. stdout_style: set the style for the screen output
3. decorate_output: add slightly more human-friendly information to output file
4. stdout_during_burnin: output progress to the screen during the burnin
5. debug: set the level of debugging output
6. output_topology_list: specify where the topology list is to be output, if anywhere
4. Controlling cBrother Priors
1. hky_kappa_prior: set the prior distribution on the HKY transition/transversion ratio.
2. hky_kappa_prior_max?: when the HKY transition/transversion ratio prior is uniform, this is the prior maximum.
3. hky_kappa_prior_mean?: when the HKY transition/transversion ratio prior is exponential, this is the prior mean.
4. hky_kappa_prior_log_mean?: when the HKY transition/transversion ratio prior is lognormal, this is the prior log mean.
5. hky_kappa_prior_variance?: when the HKY transition/transversion ratio prior is lognormal, this is the prior log variance.
6. hky_kappa_hyper_prior?: specify whether the HKY transition/transversion hyperprior parameters are estimated or fixed.
7. integrated_branch_prior: set the prior distribution on the HKY transition/transversion ratio.
8. integrated_branch_prior_mean?: when the integrated branch prior is exponential, this is the prior mean.
9. integrated_branch_prior_log_mean?: when the integrated branch prior is lognormal, this is the prior log mean.
10. integrated_branch_prior_variance?: when the integrated branch prior is lognormal, this is the prior log variance.
11. integrated_branch_hyper_prior?: specify whether the integrated branch hyperprior parameters are estimated or fixed.
12. prior_expected_parameter_chgpts: prior expected number of parameter changepoints
13. tau_prior: choice of prior distribution on the topologies along the alignment
14. tau_dim_prior: choice of prior distribution on the number of topology changepoints
15. monophyly_probability: the prior probability of monophyly throughout the alignment under the monophyly prior
16. single_event_probability: the prior probability of a single recombination event at a shared inter-genotype topology changepoint
17. prior_expected_cops: the prior expected number of topology changepoints (recombination points)
18. prior_expected_ancestry_cops: the prior expected number of topology change points where the ancestry of at least one query changes
19. prior_expected_split_ancestry_cops: the prior expected number of topology change points where the ancestry of at least one query changes and bordered by a split ancestry topology on the left or right, but not both sides.
20. prior_expected_other_cops: the prior expected number of topology change points not falling in other categories
21. lambda?: the prior expected number of change points under the single changepoint model
5. Controlling cBrother proposal distributions
1. hky_kappa_proposal: the proposal distribution for the transition/transversion ratio
2. hky_kappa_proposal_sd?: the standard deviation of the normal proposal distribution for transition/transversion ratio
3. hky_kappa_proposal_tuner?: the tuning parameter for the "exponential" proposal distribution for transition/transversion ratio
4. integrated_branch_proposal_tuner?: the tuning parameter for the "exponential" proposal distribution for integrated branch length
5. window_length: for updating changepoints, new locations are proposed uniformly within a window of this size centered on the current changepoint location
6. sigma_alpha: the standard deviation of the perturbation of transition/transversion ratios from the current transition/transversion ratio when proposing a new parameter changepoint
7. sigma_mu: the standard deviation of the perturbation of integrated branch lengths from the current integrated branch length when proposing a new parameter changepoint
8. reversible_jump_probability: tuning parameter for move probabilities; less time is dedicated to trans-dimensional moves as this number drops
9. tau_update: set the type of proposal distribution used for updating topologies
10. same_ancestry_probability: probability of proposing a new topology with the same ancestry as the existing topology
11. same_morphology_probability: probability of proposing a new topology with the same morphology (monophyly/polyphyly) as the existing topology
12. split_ancestry_topology: probability of proposing a split topology when proposing a new topology
6. Controlling the initial state of the cBrother sampler
1. initial_hky_kappa: starting transition/transversion ratio
2. initial_hky_kappa_max: when using a uniform initial distribution for transition/transversion ratio, this is the maximum
3. initial_integrated_branch: starting value for the integrated branch
7. Simulating data with cBrother: there are many options for simulating data according to the cBrother models, but there is no documentation  yet
8. Obscure commands you don't want to use.  I'm warning you!!
1. compute_likelihood: compute the likelihood during MCMC
2. random: choose the random number generator
7. Command Line Options: if links are not included, the description here should suffice
1. --continue: set to continue a previous MCMC run
2. --length: set the length of the MCMC run
3. --burnin: set the length of the MCMC burnin
4. --subsample: set the subsample rate for the MCMC run
5. --quiet?: no superfluous output
6. --dont_run?: don't actually run, just parse the command line and simulate or read the alignment data.