dosing_normalization

Last updated: 2020-06-30

Checks: 7 0

Knit directory: duplex_sequencing_screen/

This reproducible R Markdown analysis was created with workflowr (version 1.6.2). The Checks tab describes the reproducibility checks that were applied when the results were created. The Past versions tab lists the development history.

R Markdown file: up-to-date

Great! Since the R Markdown file has been committed to the Git repository, you know the exact version of the code that produced these results.

Environment: empty

Great job! The global environment was empty. Objects defined in the global environment can affect the analysis in your R Markdown file in unknown ways. For reproduciblity it’s best to always run the code in an empty environment.

Seed: set.seed(20200402)

The command set.seed(20200402) was run prior to running the code in the R Markdown file. Setting a seed ensures that any results that rely on randomness, e.g. subsampling or permutations, are reproducible.

Session information: recorded

Great job! Recording the operating system, R version, and package versions is critical for reproducibility.

Cache: none

Nice! There were no cached chunks for this analysis, so you can be confident that you successfully produced the results during this run.

File paths: relative

Great job! Using relative paths to the files within your workflowr project makes it easier to run your code on other machines.

Repository version: da11e69

Great! You are using Git for version control. Tracking code development and connecting the code version to the results is critical for reproducibility.

The results in this page were generated with repository version da11e69. See the Past versions tab to see a history of the changes made to the R Markdown and HTML files.

Note that you need to be careful to ensure that all relevant files for the analysis have been committed to Git prior to generating the results (you can use wflow_publish or wflow_git_commit). workflowr only checks the R Markdown file, but you know if there are other scripts or data files that it depends on. Below is the status of the Git repository when the results were generated:


Ignored files:
    Ignored:    .Rhistory
    Ignored:    .Rproj.user/

Untracked files:
    Untracked:  analysis/bcrabl_hill_ic50s.csv
    Untracked:  analysis/column_definitions_for_twinstrand_data_06062020.csv
    Untracked:  analysis/dose_response_curve_fitting_with_errorbars.Rmd
    Untracked:  analysis/multinomial_sims.Rmd
    Untracked:  analysis/simple_data_generation.Rmd
    Untracked:  analysis/twinstrand_growthrates_simple.csv
    Untracked:  analysis/twinstrand_maf_merge_simple.csv
    Untracked:  analysis/wildtype_growthrates_sequenced.csv
    Untracked:  code/microvariation.normalizer.R
    Untracked:  data/Combined_data_frame_IC_Mutprob_abundance.csv
    Untracked:  data/IC50HeatMap.csv
    Untracked:  data/Twinstrand/
    Untracked:  data/gfpenrichmentdata.csv
    Untracked:  data/heatmap_concat_data.csv
    Untracked:  figures_archive/
    Untracked:  output/archive/
    Untracked:  output/bmes_abstract_51220.pdf
    Untracked:  output/clinicalabundancepredictions_BMES_abstract_51320.pdf
    Untracked:  output/clinicalabundancepredictions_BMES_abstract_52020.pdf
    Untracked:  output/enrichment_simulations_3mutants_52020.pdf
    Untracked:  output/grant_fig.pdf
    Untracked:  output/grant_fig_v2.pdf
    Untracked:  output/grant_fig_v2updated.pdf
    Untracked:  output/ic50data_all_conc.csv
    Untracked:  output/ic50data_all_confidence_intervals_individual_logistic_fits.csv
    Untracked:  output/ic50data_all_confidence_intervals_raw_data.csv
    Untracked:  shinyapp/

Unstaged changes:
    Modified:   analysis/4_7_20_update.Rmd
    Modified:   analysis/E255K_alphas_figure.Rmd
    Modified:   analysis/clinical_abundance_predictions.Rmd
    Modified:   analysis/dose_response_curve_fitting.Rmd
    Modified:   analysis/misc.Rmd
    Modified:   analysis/nonlinear_growth_analysis.Rmd
    Modified:   analysis/spikeins_depthofcoverages.Rmd
    Modified:   analysis/twinstrand_spikeins_data_generation.Rmd
    Deleted:    data/README.md
    Modified:   output/twinstrand_maf_merge.csv
    Modified:   output/twinstrand_simple_melt_merge.csv

Note that any generated files, e.g. HTML, png, CSS, etc., are not included in this status report because it is ok for generated content to have uncommitted changes.

These are the previous versions of the repository in which changes were made to the R Markdown (analysis/dosing_normalization.Rmd) and HTML (docs/dosing_normalization.html) files. If you’ve configured a remote Git repository (see ?wflow_git_remote), click on the hyperlinks in the table below to view the files as they were in that past version.

File	Version	Author	Date	Message
html	da11e69	haiderinam	2020-06-30	Build site.
Rmd	278a039	haiderinam	2020-06-30	added an MCMC method for normalization. Converted normalization into a
html	3e2a9ab	haiderinam	2020-06-02	Build site.
html	32f999a	haiderinam	2020-06-02	Build site.
Rmd	d6a6fd2	haiderinam	2020-06-02	wflow_publish(“analysis/dosing_normalization.Rmd”)

This is a strategy that uses the observed GFP growth rates to estimate the approximate experimental dose and then normalizes all mutants to behave at that theoretical dose given an arbitrary hill coefficient in a dose response curve.

Theoretically, how much can an error in dosing change the replicate to replicate heterogeneity observed in an experiment?

#Looking at data that is off.
#First, I will look at data from IC50 predictions and make a correlation plot of netgr of mutants expected at 625nM vs 1.25uM.
#Next, I will look at whether any of our replicates seemed off
    a=ic50data_long%>%filter(conc%in%c(.6,1.2))
    a_cast=dcast(a,mutant~conc)

Using netgr_pred as value column: use value.var to override.

# library(reshape2)
ic50data_cast=dcast(ic50data_long,mutant~conc)

Using netgr_pred as value column: use value.var to override.

ic50data_cast6=ic50data_cast%>%dplyr::select(mutant,`0.9`)
a=ic50data_long%>%filter(conc==0.9)
# ic50data_long$`0.6`=ic50data_long%>%filter(conc==0.6)%>%dplyr::select(netgr_pred)
a=merge(ic50data_cast6,ic50data_long,by="mutant")
# ic50data_long$`0.6`=ic50data_cast$`0.6`
# ic50data_long=ic50data_long%>%mutate(`0.6`=ic50data_cast$`0.6`)
a=a%>%filter(conc%in%c(.4,.6,.8,1,1.2,1.4))
getPalette = colorRampPalette(brewer.pal(6, "Spectral"))
plotly=ggplot(a,aes(x=`0.9`,y=netgr_pred))+
  geom_abline()+
  geom_point(color="black",shape=21,size=3,aes(fill=factor(conc)))+
  scale_fill_manual(values = getPalette(length(unique(a$conc))))+
  cleanup
ggplotly(plotly)

########Making figure with predicted growth over various concentrations
ic50data_cast=dcast(ic50data_long,mutant~conc)

Using netgr_pred as value column: use value.var to override.

ic50data_cast6=ic50data_cast%>%dplyr::select(mutant,`0.8`)
a=merge(ic50data_cast6,ic50data_long,by="mutant")
a=a%>%filter(conc%in%c(.4,.8,1.2))

####Sorting mutants from more to less resistant
ic50data_long_625=a%>%filter(conc==.8)
a$mutant=factor(a$mutant,levels = as.character(ic50data_long_625$mutant[order((ic50data_long_625$y),decreasing = T)]))
###
getPalette = colorRampPalette(brewer.pal(3, "Spectral"))
ggplot(a,aes(x=`0.8`,y=netgr_pred))+
  geom_abline()+
  geom_point(color="black",shape=21,size=4,aes(fill=factor(conc)))+
  scale_x_continuous(limits=c(-.01,.06),name="Theoretical Growth rate")+
  scale_y_continuous(limits=c(-.01,.06),name="Theoretical Growth rate with 50% error in dose")+
  scale_fill_manual(values = getPalette(length(unique(a$conc))))+
  cleanup+
  theme(legend.position = "none")

Version	Author	Date
da11e69	haiderinam	2020-06-30
32f999a	haiderinam	2020-06-02

# ggsave("dosing_error_r1r2.pdf",width=3,height=3,units="in",useDingbats=F)

ggplot(a,aes(x=mutant,y=netgr_pred,color=factor(conc)))+
  geom_abline()+
  geom_point(color="black",shape=21,size=4,aes(fill=factor(conc)))+
  # scale_x_continuous(limits=c(1.33/72,1.405/72),name="Theoretical Growth rate \nat intended concentration")+
  scale_y_continuous(limits=c(-.01,.06),name="Theoretical Growth rate")+
  scale_fill_manual(values = getPalette(length(unique(a$conc))))+
  cleanup+
  theme(legend.position = "none",
        axis.title.x = element_blank(),
        axis.text.x=element_text(angle=90,hjust=.5,vjust=.5))

Version	Author	Date
da11e69	haiderinam	2020-06-30
32f999a	haiderinam	2020-06-02

# ggsave("dosing_error_r1mutant.pdf",width=5,height=3,units="in",useDingbats=F)

What Hill Coefficient to Use When inferring growth rates based on apparent dose?

We’ll need to assume a hill coefficient if we want to normalize the dose response of these mutants. Below, you will see that the typical hill coefficient in our dataset is centered around 2-3. Therefore, it is safe to assume that the hill coefficient for all the mutants is around the one forthe standard (E255K)

ic50data=read.csv("data/heatmap_concat_data.csv",header = T,stringsAsFactors = F)
# ic50data=read.csv("../data/heatmap_concat_data.csv",header = T,stringsAsFactors = F)
ic50data=ic50data[c(1:10),]
ic50data_long=melt(ic50data,id.vars = "conc",variable.name = "species",value.name = "y")
#Removing useless mutants (for example keeping only maxipreps and removing low growth rate mutants)
ic50data_long=ic50data_long%>%filter(species%in%c("Wt","V299L_H","E355A","D276G_maxi","H396R","F317L","F359I","E459K","G250E","F359C","F359V","M351T","L248V","E355G_maxi","Q252H_maxi","Y253F","F486S_maxi","H396P_maxi","E255K","Y253H","T315I","E255V"))
#Making standardized names
ic50data_long$mutant=ic50data_long$species
ic50data_long=ic50data_long%>%
  # filter(conc=="0.625")%>%
  # filter(conc=="1.25")%>%
  mutate(mutant=case_when(species=="F486S_maxi"~"F486S",
                          species=="H396P_maxi"~"H396P",
                          species=="Q252H_maxi"~"Q252H",
                          species=="E355G_maxi"~"E355G",
                          species=="D276G_maxi"~"D276G",
                          species=="V299L_H" ~ "V299L",
                          species==mutant ~as.character(mutant)))

# ic50data_long_625$species[order((ic50data_long_625$y),decreasing = T)]

#In the next step, I'm ordering mutants by decreasing resposne to the 625nM dose. Then I use this to change the levels of the species factor from more to less resistant. This helps with ggplot because now I can color the mutants with decreasing resistance
ic50data_long_625=ic50data_long%>%filter(conc==.625)
ic50data_long$species=factor(ic50data_long$species,levels = as.character(ic50data_long_625$species[order((ic50data_long_625$y),decreasing = T)]))
ic50data_long$mutant=factor(ic50data_long$mutant,levels = as.character(ic50data_long_625$mutant[order((ic50data_long_625$y),decreasing = T)]))

########Four parameter logistic########
#Reference: https://journals.plos.org/plosone/article/file?type=supplementary&id=info:doi/10.1371/journal.pone.0146021.s001
#In short: For each dose in each species, get the response
# rm(list=ls())
ic50data_long_model=data.frame()
# hill_coefficients=data.frame()
hill_coefficients_cum=data.frame()
ic50data_long$species=ic50data_long$mutant
for (species_curr in sort(unique(ic50data_long$species))){
  ic50data_species_specific=ic50data_long%>%filter(species==species_curr)
  x=ic50data_species_specific$conc
  y=ic50data_species_specific$y
  #Next: Appproximating Response from dose (inverse of the prediction)
  ic50.ll4=drm(y~conc,data=ic50data_long%>%filter(species==species_curr),fct=LL.3(fixed=c(NA,1,NA)))
    b=coef(ic50.ll4)[1]
    c=0
    d=1
    e=coef(ic50.ll4)[2]
  ###Getting predictions
  ic50data_species_specific=ic50data_species_specific%>%group_by(conc)%>%mutate(y_model=c+((d-c)/(1+exp(b*(log(conc)-log(e))))))
  ic50data_species_specific=data.frame(ic50data_species_specific) #idk why I have to end up doing this
  ic50data_long_model=rbind(ic50data_long_model,ic50data_species_specific)
  hill_coefficients_curr=data.frame(cbind(species_curr,b,e))
  # hill_coefficients$species=species_curr
  # hill_coefficients$hill=b
  hill_coefficients_cum=rbind(hill_coefficients_cum,hill_coefficients_curr)
}
ic50data_long=ic50data_long_model

#In the next step, I'm ordering mutants by decreasing resposne to the 625nM dose. Then I use this to change the levels of the species factor from more to less resistant. This helps with ggplot because now I can color the mutants with decreasing resistance
ic50data_long_625=ic50data_long%>%filter(conc==.625)
ic50data_long$species=factor(ic50data_long$species,levels = as.character(ic50data_long_625$species[order((ic50data_long_625$y_model),decreasing = T)]))

#Adding drug effect
##########Changed this on 2/20. Using y from 4 parameter logistic rather than raw values
ic50data_long=ic50data_long%>%
  filter(!species=="Wt")%>%
  mutate(drug_effect=-log(y_model)/72)

#Adding Net growth rate
ic50data_long$netgr_pred=.05-ic50data_long$drug_effect

hill_coefficients_cum$b=as.numeric(hill_coefficients_cum$b)
ggplot(hill_coefficients_cum,aes(b))+geom_histogram()

`stat_bin()` using `bins = 30`. Pick better value with `binwidth`.

Version	Author	Date
da11e69	haiderinam	2020-06-30

ggplot(hill_coefficients_cum,aes(x=species_curr,y=b))+geom_point()

Version	Author	Date
da11e69	haiderinam	2020-06-30

#Maybe we can assume a hill of 2?



###Preparing spreadsheet of hill coefficients for MCMC. 06072020
a=hill_coefficients_cum%>%dplyr::select(mutant=species_curr,hill_coef=b,ic50=e)
rownames(a)=NULL
# write.csv(a,"bcrabl_hill_ic50s.csv")