R/dNdS_segment.R

In vivaldi: Viral Variant Location and Diversity

#' dNdS_segment
#'
#' Reads in a dataframe that has been arranged (arrange_data), filtered (filter_variants), and annotated (prepare_annotations), calculates dNdS, and outputs plots
#'
#' @name dNdS_segment
#' @param annotation_df A rearranged, filtered, and annotated vcf dataframe - must be for amino-acid specific calculations, cannot be the same as the dataframe used for SNP calculations
#' @param SPLICEFORMS A character vector of isoform names
#' @return A plot showing the dN/dS ratio for each splice form (rather than segment) for each sample
#' @export
#' @examples
#' # Sample Data
#' head(example_filtered_SNV_df)
#' dim(example_filtered_SNV_df)
#'
#' # Plot showing the dN/dS ratio for each splice form
#' SPLICEFORMS = c("H1N1_PB2.1", "H1N1_PB1.1","H1N1_NS.2")
#' dNdS_segment(example_filtered_SNV_df, SPLICEFORMS)
dNdS_segment = function(annotation_df, SPLICEFORMS){

  annotation_df$feature_id = factor(annotation_df$feature_id, levels = SPLICEFORMS)

  df_dN = dplyr::filter(annotation_df, annotation == "missense_variant" | annotation == "nonsense_variant") %>%
    dplyr::group_by(sample, feature_id) %>% droplevels() %>% dplyr::tally()
  df_dS = dplyr::filter(annotation_df, annotation == "synonymous_variant") %>%
    dplyr::group_by(sample, feature_id) %>% droplevels() %>% dplyr::tally()

  df = merge(df_dN, df_dS,  by = c("sample","feature_id")) %>% dplyr::select(sample,feature_id,n.x,n.y)
  df$dNdS = (df$n.x) / (df$n.y)

  plot = ggplot2::ggplot(df,ggplot2::aes(x = sample , y = dNdS)) +
    ggplot2::geom_point() +
    ggplot2::facet_grid(~feature_id) +
    ggplot2::ggtitle("dNdS Ratio Per Splice Form") +
    ggplot2::theme(legend.key = ggplot2::element_blank(),
          strip.background = ggplot2::element_rect(colour="black", fill="white"),
          axis.text.x = ggplot2::element_text(angle = 90, vjust = 0.5, hjust=1))

  print(plot)
}