run_SMC: Wrapper function for the Stratification of a Mutational...
In eilslabs/YAPSA: Yet Another Package for Signature Analysis

Description Usage Arguments Value See Also Examples

run_SMC takes as input a big dataframe constructed from a vcf-like file of a whole cohort. This wrapper function calls custom functions to construct a mutational catalogue and stratify it according to categories indicated by a special column in the input dataframe:

create_mutation_catalogue_from_df
adjust_number_of_columns_in_list_of_catalogues

This stratification yields a collection of stratified mutational catalogues, these are reformatted and sent to the custom function SMC and thus indirectly to LCD_SMC to perform a signature analysis of the stratified mutational catalogues. The result is then handed over to plot_SMC for visualization.

run_SMC(my_table, this_signatures_df, this_signatures_ind_df, this_subgroups_df,
  column_name, refGenome, cohort_method_flag = "all_PIDs",
  in_strata_order_ind = seq_len(length(unique(my_table[, column_name]))),
  wordLength = 3, verbose_flag = 1, target_dir = NULL,
  strata_dir = NULL, output_path = NULL, in_all_exposures_df = NULL,
  in_rownames = c(), in_norms = NULL, in_label_orientation = "turn",
  this_sum_ind = NULL)

`my_table`	A big dataframe constructed from a vcf-like file of a whole cohort. The first columns are those of a standard vcf file, followed by an arbitrary number of custom or user defined columns. One of these must carry a PID (patient or sample identifyier) and one must be the category used for stratification.
`this_signatures_df`	A numeric data frame `W` in with `n` rows and `l` columns, `n` being the number of features and `l` being the number of signatures
`this_signatures_ind_df`	A data frame containing meta information about the signatures
`this_subgroups_df`	A data frame indicating which PID (patient or sample identifyier) belongs to which subgroup
`column_name`	Name of the column in `my_table` which is going to be used for stratification
`refGenome`	FaFile of the reference genome to extract the motif context of the variants in `my_table`
`cohort_method_flag`	Either or several of `c("all_PIDs","cohort","norm_PIDs")`, representing alternative ways to average over the cohort.
`in_strata_order_ind`	Index vector defining reordering of the strata
`wordLength`	Integer number defining the length of the features or motifs, e.g. 3 for tripletts or 5 for pentamers
`verbose_flag`	Verbose if `verbose_flag=1`
`target_dir`	Path to directory where the results of the stratification procedure are going to be stored if non-NULL.
`strata_dir`	Path to directory where the mutational catalogues of the different strata are going to be stored if non-NULL
`output_path`	Path to directory where the results, especially the figures produced by `plot_SMC` are going to be stored.
`in_all_exposures_df`	Optional argument, if specified, `H`, i.e. the overall exposures without stratification, is set to equal `in_all_exposures_df`. This is equivalent to forcing the LCD_SMC procedure to use e.g. the exposures of a previously performed NMF decomposition.
`in_rownames`	Optional parameter to specify rownames of the mutational catalogue `V` i.e. the names of the features.
`in_norms`	If specified, vector of the correction factors for every motif due to differing trinucleotide content. If null, no correction is applied.
`in_label_orientation`	Whether or not to turn the labels on the x-axis.
`this_sum_ind`	Optional set of indices for reordering the PIDs

A list with entries exposures_list, catalogues_list, cohort and name_list.

exposures_list: The list of s strata specific exposures Hi, all are numerical data frames with l rows and m columns, l being the number of signatures and m being the number of samples
catalogues_list: A list of s strata specific cohortwide (i.e. averaged over cohort) normalized exposures
cohort: subgroups_df adjusted for plotting
name_list: Names of the contructed strata.

create_mutation_catalogue_from_df

normalizeMotifs_otherRownames

plot_SMC

 library(BSgenome.Hsapiens.UCSC.hg19)
 data(sigs)
 data(lymphoma_test)
 data(lymphoma_cohort_LCD_results)
 strata_list <- 
   cut_breaks_as_intervals(lymphoma_test_df$random_norm,
                           in_outlier_cutoffs=c(-4,4),
                           in_cutoff_ranges_list=list(c(-2.5,-1.5),
                                                      c(0.5,1.5)),
                           in_labels=c("small","intermediate","big"))
 lymphoma_test_df$random_cat <- strata_list$category_vector
 choice_ind <- (names(lymphoma_Nature2013_COSMIC_cutoff_exposures_df) 
                %in% unique(lymphoma_test_df$PID))
 lymphoma_test_exposures_df <- 
   lymphoma_Nature2013_COSMIC_cutoff_exposures_df[,choice_ind]
 temp_subgroups_df <- make_subgroups_df(lymphoma_test_df,
                                        lymphoma_test_exposures_df)
 mut_density_list <- run_SMC(lymphoma_test_df,
                             AlexCosmicValid_sig_df,
                             AlexCosmicValid_sigInd_df,
                             temp_subgroups_df,
                             column_name="random_cat",
                             refGenome=BSgenome.Hsapiens.UCSC.hg19,
                             cohort_method_flag="norm_PIDs",
                             in_rownames = rownames(AlexCosmicValid_sig_df))