A Versatile Procedure to Generate Genome-Wide Spatiotemporal Program of Replication in Yeast Species

Abstract

Here, we describe a complete protocol, comprising both the experimental and the analytical procedures, that allows to generate genome-wide spatiotemporal program of replication and to find the location of chromosomally active replication origins in yeast. The first step consists on synchronizing a cell population by physical discrimination of G1 cells according to their sedimentation coefficient. G1 cells are then synchronously released into S-phase and time-point samples are regularly taken until they reach the G2 phase. Progression through the cell cycle is monitored by measuring DNA content variation by flow cytometry. DNA samples, covering the entire S-phase, are then extracted and analyzed using deep sequencing. The gradual change of DNA copy number is measured to determine the mean replication time along the genome. A simple method of peak calling allows to infer from the replication profile the location of replication origins along the chromosomes. Our protocol is versatile enough to be applied to virtually any yeast species of interest and generate its replication profile.

Appendix

List of commands for BWA:

    ‘genome indexing

        ./bwa index genome.fasta

    ‘read mapping

        ./bwa aln -n 0 -o 0 genome.fasta my_reads.fastq > results_temp.sai

    ‘format as sam file

        ./bwa samse -n 0 -f results.sam genome.fasta results_temps.sai my_reads.fastq

    ‘filter the results.sam file

        awk ‘{if($5==37 && $12==”XT:A:U”)  print}’ results.sam > results_filtered.sam

R script for Trep estimation:

       ‘Prepare one Input table for each chromosome. The first column named “ID” contains a unique ‘ID for each 500 window and the following columns (one column per time point) contain the ‘Rscaled ratios. The following script is given for 8 times points j = [25, 30, 35, 40, 45, 50, 55, 60] ‘and the corresponding columns containing the Rscaled ratio are named T1 to T8.

      t = read.table("Input table", header=TRUE)

      lev = levels(t$ID)

      lev = lev[2:length(lev)]

      tab = matrix(nrow = length(lev), ncol = 3)

      for(i in 1:length(lev))

      {

              if(lev[i] != "")

              {

                  print(paste("Sonde ", as.character(i), " / ", as.character(length(lev)), " : ",

      lev[i], sep = ""))

      ‘format the data.

                  me = seq(1,8)

      me[1] = t$T1_Adj[t$ID == lev[i]]

      me[2] = t$T2_Adj[t$ID == lev[i]]

      me[3] = t$T3_Adj[t$ID == lev[i]]

      me[4] = t$T4_Adj[t$ID == lev[i]]

      me[5] = t$T5_Adj[t$ID == lev[i]]

      me[6] = t$T6_Adj[t$ID == lev[i]]

      me[7] = t$T7_Adj[t$ID == lev[i]]

      me[8] = t$T8_Adj[t$ID == lev[i]]

      ‘Here enter the list of values for j manually

                  time=c(25,30,35,40,45,50,55,60)

      ‘Loess fitting of scaled data for each 500 bp window

                  l = loess(me~time, span = 0.75, family = "gaussian")

      ‘Estimation of time (Trep) when the Rscaled ratio is equal to 1.5.

                  pred = -1

      ‘Enter inf value manually (inf value = min(j))

                  inf = 25

      ‘Enter sup value manually (sup value = max(j))

                  sup = 60

                  lim = 0.03

                  while(pred == -1)

                  {

                      temps = mean(c(inf, sup))

                      if(sup - inf < 0.03)

                      {

                          pred = predict(l, temps)

                          break

                      }

                      if(predict(l, temps) > 1.5)

                          sup = temps

                      else if(predict(l, temps) < 1.5)

                          inf = temps

                      else

                          pred = predict(l, temps)

                  }

                  print(paste("Point ", as.character(pred), " trouve a ", as.character(temps),

      sep = ""))

                  tab[i, ] = c(lev[i], temps, pred)

              }

            }

      colnames(tab) = c("Probe", "Trep", "Rscaled")

      write.table(tab, "My_results.txt", row.names = FALSE, sep = "\t", quote = FALSE)

R script for spatiotemporal replication profile:

      ‘Prepare one Input table for each chromosome. Each input table contains 3 columns: ‘Chromosome, Position, Trep. Position are expressed in base pair.

      t = read.table("Input table", header = TRUE)

      threshold = 60000

      sp = threshold/max(t$Position)

      l = loess(t$Trep~t$Position, span = sp)

      tabresult = matrix(nrow=l$n, ncol=3, data=c(l$x, l$y, l$fitted))

      colnames(tabresult) = c("Position", "Trep", "Tfit")

      Result = as.data.frame(tabresult)

      write.table(Result, file = "profile.txt",row.names = FALSE, sep=",")