Theoretical and sequenced composition of the Zymo DNA standard

Zymo Research have a nice DNA standard for microbiomics, which can be used as a control in 16S sequencing experiments. I wanted to know how good the sequenced and theoretical composition correlate. Here is data from a recent MiSeq run (at the NCCT, where I work now), with the ZymoBIOMICS DNA standard used as a control.

---

---

Pretty good, right?
Next, I will post data from a metagenomics run with this standard mix.

Code

For reproducibility, the code below was used. In short, the fastq reads were mapped (minimap2) against a database (available here) containing the 16S rDNA genes of the 8 organisms in the standard. The resulting PAF file was used directly to count the number of reads mapping to each taxon (keeping only alignments > 250 bp, the read length was 300 bp). The theoretical composition of the Zymo DNA standard was taken from the manual.

seqtk seq -l 80 ssdb.fas > ssdb1.fas # prepare the fasta file so that it can be read by minimap2
minimap2 -x sr ssdb1.fas \
ZymoControl_S86_R1_001.fastq.gz  \
ZymoControl_S86_R2_001.fastq.gz > zymo.paf # run the mapping

# in R:
library(data.table)
library(tidyverse)
df <- fread("zymo.paf") %>% 
        filter(V10 > 250, V12 == 0) %>% # take only good alignments
        mutate(species = str_extract(V6, "[A-z]+_[a-z]+")) # clean species name

miseq <- df %>% 
          group_by(species) %>% 
          summarise(n = n(), miseq_percent = n/nrow(.)*100)

zymo <- fread("Zymo_control_table.txt") %>% mutate(species = str_c(V1, V2, sep = "_")) # read Zymo composition data
final <- zymo %>% # make final df
          select(species, zymo_percent = V4) %>% 
          left_join(miseq, by = "species")

# ...plot
final %>% 
  #arrange(desc(percent)) %>% #bug!! if I do this, the values get mixed up!! the factor trick is also not working
  apex(mapping = aes(source, round(percent,2), fill = species), type = "bar", 
       width = 800, height = 300) %>% 
  ax_chart(stacked = TRUE, stackType = "normal") %>%
  ax_legend(horizontalAlign = "left", position = "right", 
            itemMargin = list(horizontal =5, vertical = 0)) %>%
  ax_colors(scales::brewer_pal(palette = "Set2")(8)) %>%
  ax_dataLabels(enabled = TRUE) %>%
  ax_yaxis(max = 100) %>%
  ax_labs(title = "Theoretical (zymo) and measured (MiSeq) composition of the Zymo standard",
          subtitle = "MiSeq run (2 x 300) of amplified (20 PCR cycles) V3-V4 region, 359 191 read pairs")

Disclaimer

I am not affiliated with Zymo Research in any way.