Selects the best model (mu, rho) combination based on total log-likelihood and distance to integer copy number, refines segment calls, and writes results to output files.
Usage
SelectFinalModel(
results,
top_likelihood_rows,
groupinfo,
prefix,
gender,
out_dir,
model_source,
seg,
callcov,
modelminAIsize,
modelminprobes,
minsf
)Arguments
- results
Data frame or tiblle. Likelihood results calculated by RunCallikelihood function.
- top_likelihood_rows
Data frame or tibble. Top likelihood calls for each segment, including columns
mu,rho,log_MAF_likelihood,major,minor,Tag, etc.- groupinfo
Data frame or tibble. Information for estimating minimum purity (see
EstimateMinPurity).- prefix
Character. Output file prefix.
- gender
Character. Sample gender ("male" or "female").
- out_dir
Character. Output directory for saving results.
- model_source
Character. Description of model source (e.g., "Coverage", "Coverage + MAF").
- seg
Data frame or tibble. Segment information.
- callcov
Numeric. Subclonal events calling cutoff based on coverage.
- modelminAIsize
Numeric. Minimum segment size for model inclusion.
- modelminprobes
Integer. Minimum number of probes for model inclusion.
- minsf
Numeric. Minimum scale factor will be estimated. default is 0.4. range 0~0.4.
Value
A list with:
- Final_model
The final selected model (see
ClusterModels).- models
The full model table (before and after refinement).
- refined_calls
The final calls of tier1 models.