Skip to content

Commit

Permalink
add params description of mobster/viber/pyclone + modify test
Browse files Browse the repository at this point in the history
  • Loading branch information
@valerianilucrezia
valerianilucrezia committed Jan 10, 2025
1 parent 5844459 commit 737bebf
Show file tree
Hide file tree
Showing 3 changed files with 83 additions and 49 deletions.
10 changes: 4 additions & 6 deletions conf/test.config
View file
Original file line number Diff line number Diff line change
Expand Up @@ -42,18 +42,16 @@ params {
vep_cache_version = "110"
vep_species = "Homo_sapiens"

mobster_K = 2
mobster_maxIter = 100
mobster_samples = 2
pyclonevi_n_restarts = 50

sparsesignatures_K = "2:3"
sparsesignatures_nmf_runs = "3"
sparsesignatures_iterations = "10"
sparsesignatures_max_iterations_lasso = "100"
sparsesignatures_cross_validation_repetitions = "10"
sparsesignatures_max_iterations_lasso = "80"
sparsesignatures_cross_validation_repetitions = "8"
sparsesignatures_cross_validation_iterations = "5"
sparsesignatures_lambda_values_alpha = "c(0.00, 0.01)"
sparsesignatures_lambda_values_beta = "c(0.01, 0.05)"

tools = "mobster,viber,pyclone-vi,sparsesignatures"
tools = "pyclone-vi,sparsesignatures"
}
114 changes: 75 additions & 39 deletions nextflow_schema.json
View file
Original file line number Diff line number Diff line change
Expand Up @@ -50,15 +50,15 @@
"default": "mobster,viber,pyclone-vi,sparsesignatures,sigprofiler",
"description": "List of tools for running the pipeline."
},
"publish_dir_mode": {
"type": "string",
"default": "copy",
"description": "Method used to save pipeline results to output directory."
},
"filter": {
"type": "boolean",
"default": true,
"description": "Flag for filtering or not QC mutations."
},
"publish_dir_mode": {
"type": "string",
"default": "copy",
"description": "Method used to save pipeline results to output directory."
}
},
"required": ["genome"],
Expand Down Expand Up @@ -180,7 +180,8 @@
},
"cnaqc_kernel_adjust": {
"type": "integer",
"default": 1
"default": 1,
"description": "For KDE-based matches the adjust density parameter; see density."
},
"cnaqc_matching_strategy": {
"type": "string",
Expand Down Expand Up @@ -250,19 +251,27 @@
"properties": {
"pyclonevi_density": {
"type": "string",
"default": "beta-binomial"
"default": "beta-binomial",
"description": "The probability density used to model the read count data. Choices are beta-binomial and binomial.",
"help_text": "binomial is a common choice for sequencing data. beta-binomial is useful when the data is over-dispersed which has been observed frequently in sequencing data."
},
"pyclonevi_n_restarts": {
"type": "integer",
"default": 100
"default": 100,
"description": "Number of random restarts of variational inference.",
"help_text": "More restarts will have a higher probability of finding the optimal variational approximation. This also increases running time."
},
"pyclonevi_n_grid_point": {
"type": "integer",
"default": 100
"default": 100,
"description": "Number of grid points used for approximating the posterior distribution.",
"help_text": " Higher values should be used for deeply sequenced data."
},
"pyclonevi_n_cluster": {
"type": "integer",
"default": 20
"default": 20,
"description": "The number of clusters to use while fitting.",
"help_text": "This should be set to a value larger than the expected number of clusters. The software will then automatically determine how many to use. In general this value should increase if as more samples are used."
}
},
"fa_icon": "fas fa-briefcase"
Expand All @@ -275,63 +284,77 @@
"properties": {
"mobster_K": {
"type": "string",
"default": "1:5"
"default": "1:5",
"description": "A vector with the number of Beta components to use. All values of K must be positive and strictly greater than 0."
},
"mobster_init": {
"type": "string",
"default": "peaks"
"default": "peaks",
"description": "Initial values for the parameters of the model. Can be either \"random\" or \"peaks\" ."
},
"mobster_samples": {
"type": "integer",
"default": 5
"default": 5,
"description": "Number of fits that should be attempted for each configuration of the model tested."
},
"mobster_tail": {
"type": "string",
"default": "c(TRUE,FALSE)"
"default": "c(TRUE,FALSE)",
"description": "Boolean value whether to use or not tail mutations for subclonal deconvolution."
},
"mobster_epsilon": {
"type": "number",
"default": 1e-10
"default": 1e-10,
"description": "Tolerance for convergency estimation."
},
"mobster_maxIter": {
"type": "integer",
"default": 250
"default": 250,
"description": "Maximum number of steps for a fit."
},
"mobster_fit_type": {
"type": "string",
"default": "MM"
"default": "MM",
"description": "A string that determines the type of fit. Can be either \"MLE\" , for the Maximum Likelihood Estimate of the Beta parameters, or \"MM\"for the Moment Matching."
},
"mobster_seed": {
"type": "integer",
"default": 12345
"default": 12345,
"description": "Seed for the random numbers generator"
},
"mobster_model_selection": {
"type": "string",
"default": "reICL"
"default": "reICL",
"description": "Score to minimize to select the best model; this has to be one of 'reICL', 'ICL', 'BIC', 'AIC' or 'NLL'."
},
"mobster_trace": {
"type": "string",
"default": "FALSE"
"default": "FALSE",
"description": "Boolean value whether to return the trace of model fit."
},
"mobster_parallel": {
"type": "string",
"default": "TRUE"
"default": "TRUE",
"description": "Boolean value whether to run the fit in parallel."
},
"mobster_pi_cutoff": {
"type": "number",
"default": 0.02
"default": 0.02,
"description": "The minimum mixing proportion of a cluster to be returned as output."
},
"mobster_n_cutoff": {
"type": "integer",
"default": 10
"default": 10,
"description": "The minimum number of mutations assigned to a cluster to be returned as output."
},
"mobster_silent": {
"type": "string",
"default": "FALSE"
},
"mobster_auto_setup": {
"type": "string",
"default": "NULL"
"default": "NULL",
"description": "Overrides all the parameters with a predefined set of values, in order to implement different analyses."
}
},
"fa_icon": "fas fa-briefcase"
Expand All @@ -344,55 +367,69 @@
"properties": {
"viber_K": {
"type": "integer",
"default": 10
"default": 10,
"description": "The maximum number of clusters returned"
},
"viber_samples": {
"type": "integer",
"default": 10
"default": 10,
"description": "The number of fits to be computed."
},
"viber_alpha_0": {
"type": "number",
"default": 1e-6
"default": 1e-6,
"description": "The concentration parameter of the Dirichlet mixture."
},
"viber_a_0": {
"type": "integer",
"default": 1
"default": 1,
"description": "The prior Beta hyperparameter for each Binomial component a"
},
"viber_b_0": {
"type": "integer",
"default": 1
"default": 1,
"description": "The prior Beta hyperparameter for each Binomial component b"
},
"viber_maxIter": {
"type": "integer",
"default": 5000
"default": 5000,
"description": "The maximum number of fit iterations"
},
"viber_epsilon_conv": {
"type": "number",
"default": 1e-10
"default": 1e-10,
"description": "The epsilon to measure convergence as ELBO absolute difference"
},
"viber_q_init": {
"type": "string",
"default": "prior"
"default": "prior",
"description": "Initialization of the q-distribution to compute the approximation of the posterior distributions.",
"help_text": "This can be set in three different waysL equal to the prior (q_init = 'prior'), via kmeans clustering (q_init = 'kmeans') and capturing points which are private to each dimension (q_init = 'private')."
},
"viber_trace": {
"type": "string",
"default": "FALSE"
"default": "FALSE",
"description": "Boolean value whether to return the trace of model fit."
},
"viber_binomial_cutoff": {
"type": "number",
"default": 0.05
"default": 0.05,
"description": "The minimum Binomial success probability when applying a heuristic procedure to filter clusters after Variational Inference."
},
"viber_pi_cutoff": {
"type": "number",
"default": 0.02
"default": 0.02,
"description": "The minimum size of the mixture component when applying a heuristic procedure to filter clusters after Variational Inference."
},
"viber_re_assign": {
"type": "string",
"default": "FALSE"
"default": "FALSE",
"description": "Boolean value whether point assigned to a cluster that is filtered our, are re-assigned from the density function."
},
"viber_dimensions_cutoff": {
"type": "integer",
"default": 1
"default": 1,
"description": "The minimum number of dimensions where we want to detect a Binomial component when applying a heuristic procedure to filter clusters after Variational Inference."
}
},
"fa_icon": "fas fa-briefcase"
Expand Down Expand Up @@ -815,6 +852,5 @@
{
"$ref": "#/$defs/other"
}
],
"properties": {}
]
}
8 changes: 4 additions & 4 deletions workflows/tumourevo.nf
View file
Original file line number Diff line number Diff line change
Expand Up @@ -65,8 +65,9 @@ main:
params.vep_cache_version,
vep_cache,
ch_extra_files)

vcf_file = FORMATTER_VCF(VCF_ANNOTATE_ENSEMBLVEP.out.vcf_tbi, "vcf")
FORMATTER_CNA(input_cna, "cna")
cna_file = FORMATTER_CNA(input_cna, "cna")

join_input = vcf_file.join(input_bam).map{ meta, rds, bam, bai ->
[ meta, rds, bam, bai ] }
Expand All @@ -76,15 +77,14 @@ main:
}

out_lifter = LIFTER(join_input.to_lift, fasta)

rds_input = join_input.multisample.map{ meta, rds, bam, bai ->
[meta, rds]
}
vcf_rds = rds_input.concat(out_lifter)

annotation = DRIVER_ANNOTATION(vcf_rds, drivers_table)
cna_out = FORMATTER_CNA.out

in_cnaqc = cna_out.join(annotation)
in_cnaqc = cna_file.join(annotation)
QC(in_cnaqc)

if (params.filter == true){
Expand Down

0 comments on commit 737bebf

Please sign in to comment.