R/nx_lx.R

In A-BN/fastaUtils: Simple fasta exploration and manipulation

#' Calculate assembly statistics N50, L50, N90, L90
#' @param contig_df, a contigs data.frame with two cols, name and size
#' @return a data.frame containing N50, L50, N90, L90 and
#' the input data.frame with cols cum_size and contig_n added
nx_lx <- function(contig_df) {
  contig_df <- contig_df[order(contig_df$size, decreasing = TRUE), ]
  tot_contig_size <- sum(contig_df$size)
  contig_df$cum_size <- cumsum(contig_df$size)
  contig_df$contig_n <- 1:nrow(contig_df)
  
  L50 <- sum((contig_df$cum_size) < (tot_contig_size * 0.5)) + 1
  N50 <- contig_df$size[L50]
  
  L90 <- sum((contig_df$cum_size) < (tot_contig_size * 0.9)) + 1
  N90 <- contig_df$size[L90]

  metrics_df <- data.frame(x = c(50, 90), n = c(N50, N90), l = c(L50, L90))
  metrics_df$cum_Nx <- contig_df$cum_size[c(L50, L90)]

  return(list(contig_df, metrics_df))
}