| Title: | Simple Metrics to Summarize Growth Curves |
|---|---|
| Description: | Fits the logistic equation to microbial growth curve data (e.g., repeated absorbance measurements taken from a plate reader over time). From this fit, a variety of metrics are provided, including the maximum growth rate, the doubling time, the carrying capacity, the area under the logistic curve, and the time to the inflection point. Method described in Sprouffske and Wagner (2016) <doi:10.1186/s12859-016-1016-7>. |
| Authors: | Kathleen Sprouffske [aut, cre] |
| Maintainer: | Kathleen Sprouffske <[email protected]> |
| License: | GPL (>= 2) |
| Version: | 0.3.1 |
| Built: | 2025-03-02 03:33:46 UTC |
| Source: | https://github.com/sprouffske/growthcurver |
This is a constructor function for the "gcfit" class. This class is most
often obtained as the return value when calling
SummarizeGrowth.
gcfit(gc_vals, log_mod, data_t, data_n)gcfit(gc_vals, log_mod, data_t, data_n)
gc_vals |
An object of class gcvals that contains the summarized
metrics from fitting the growth model to a set of
experimental observations. This is where the fitness proxy
parameters can be found. See |
log_mod |
An object of class nlsModel that contains the results of fitting the logistic growth model to the data |
data_t |
A numeric vector of times |
data_n |
A numeric vector of cell count or absorbance readings |
An object of class gcfit, which is a list of three objects, that combines the parameters (vals = gc_vals, model = log_mod, data = list(data_t, data_n))
Constructor function for the "gcvals" class. This object is most often
obtained when calling SummarizeGrowth (it is the first
parameter in the gcvals object).
gcvals( k, k_se, k_p, n0, n0_se, n0_p, r, r_se, r_p, sigma, df, t_mid, dt, auc_l, auc_e, note )gcvals( k, k_se, k_p, n0, n0_se, n0_p, r, r_se, r_p, sigma, df, t_mid, dt, auc_l, auc_e, note )
k |
The carrying capacity parameter |
k_se |
The standard error of the carrying capacity parameter |
k_p |
The p value of the carrying capacity parameter |
n0 |
The initial population size |
n0_se |
The standard error of the initial population size |
n0_p |
The p value of the initial population size |
r |
The growth rate |
r_se |
The standard error of the growth rate |
r_p |
The p value of the growthrate |
sigma |
Residual standard error from non-linear least squares fit of the model to the data |
df |
Degrees of freedom |
t_mid |
The time at the inflection point of the logistic curve (occurs at half of the carrying capacity) |
dt |
The maximum doubling time, obtained by evaluating the the unrestrained growth of the population with growth rate r |
auc_l |
The area under the curve of the fitted logistic equation from time 0 to time t |
auc_e |
The area under the curve of the measurements. |
note |
Feedback on common problems with fitting the logistic curve to the data |
An object of class gcvals.
A dataset containing absorbance measurements over time of microbes growing in a plate reader for 1 day. The growth curves for a whole plate are included.
growthdatagrowthdata
A data frame with 145 observations and 97 variables:
time, in hours
absorbance readings of well A1
absorbance readings of well A2
absorbance readings of well A3
absorbance readings of well A4
absorbance readings of well A5
absorbance readings of well A6
absorbance readings of well A7
absorbance readings of well A8
absorbance readings of well A9
absorbance readings of well A10
absorbance readings of well A11
absorbance readings of well A12
absorbance readings of well B1
absorbance readings of well B2
absorbance readings of well B3
absorbance readings of well B4
absorbance readings of well B5
absorbance readings of well B6
absorbance readings of well B7
absorbance readings of well B8
absorbance readings of well B9
absorbance readings of well B10
absorbance readings of well B11
absorbance readings of well B12
absorbance readings of well C1
absorbance readings of well C2
absorbance readings of well C3
absorbance readings of well C4
absorbance readings of well C5
absorbance readings of well C6
absorbance readings of well C7
absorbance readings of well C8
absorbance readings of well C9
absorbance readings of well C10
absorbance readings of well C11
absorbance readings of well C12
absorbance readings of well D1
absorbance readings of well D2
absorbance readings of well D3
absorbance readings of well D4
absorbance readings of well D5
absorbance readings of well D6
absorbance readings of well D7
absorbance readings of well D8
absorbance readings of well D9
absorbance readings of well D10
absorbance readings of well D11
absorbance readings of well D12
absorbance readings of well E1
absorbance readings of well E2
absorbance readings of well E3
absorbance readings of well E4
absorbance readings of well E5
absorbance readings of well E6
absorbance readings of well E7
absorbance readings of well E8
absorbance readings of well E9
absorbance readings of well E10
absorbance readings of well E11
absorbance readings of well E12
absorbance readings of well F1
absorbance readings of well F2
absorbance readings of well F3
absorbance readings of well F4
absorbance readings of well F5
absorbance readings of well F6
absorbance readings of well F7
absorbance readings of well F8
absorbance readings of well F9
absorbance readings of well F10
absorbance readings of well F11
absorbance readings of well F12
absorbance readings of well G1
absorbance readings of well G2
absorbance readings of well G3
absorbance readings of well G4
absorbance readings of well G5
absorbance readings of well G6
absorbance readings of well G7
absorbance readings of well G8
absorbance readings of well G9
absorbance readings of well G10
absorbance readings of well G11
absorbance readings of well G12
absorbance readings of well H1
absorbance readings of well H2
absorbance readings of well H3
absorbance readings of well H4
absorbance readings of well H5
absorbance readings of well H6
absorbance readings of well H7
absorbance readings of well H8
absorbance readings of well H9
absorbance readings of well H10
absorbance readings of well H11
absorbance readings of well H12
This function gives the number of cells or absorbance (N) at time t when the parameters to the logistic equation are K, N0, and r.
NAtT(k, n0, r, t)NAtT(k, n0, r, t)
k |
The carrying capacity |
n0 |
The initial population size (absorbance or individuals) |
r |
The exponential "growth rate" |
t |
The time at which you want to know N |
The number of cells, or N, at time t
This function finds the parameters that describe the input data's growth. It does so by fitting the logistic curve to your growth curve measurements.
SummarizeGrowth(data_t, data_n, t_trim = 0, bg_correct = "min", blank = NA)SummarizeGrowth(data_t, data_n, t_trim = 0, bg_correct = "min", blank = NA)
data_t |
A vector of timepoints (data_n must also be provided and be the same length). |
data_n |
A vector of cell counts or absorbance readings. |
t_trim |
Measurements taken after this time should not be included in fitting the curve. If stationary phase is variable, this may give you a better fit. A value of 0 means no trimming. Defaults to no trimming (0). |
bg_correct |
The background correction method to use. No background correction is performed for the default "none". Specifying "min" subtracts the smallest value in a column from all the rows in that column, and specifying "blank" subtracts the values from the blank vector from the data_n vector. |
blank |
A vector of absorbance readings from a blank well (typically contains only media) used for background correction. The corresponding blank value is subtracted from the data_n vector for each timepoint. Defaults to NA. |
The logistic curve equation is
where is the number
of cells (or the absorbance reading) at time t, is the initial
cell count (or absorbance reading), K is the carrying capacity, and r is the
growth rate.
The fitness proxies returned are the parameters of the logistic equation
and the area under the curve (a measure that integrates the effects
of , K, and r). See gcfit for more documentation on these.
An object of type gcfit containing the "fitness" proxies, as well as the input data and the fitted model.
See the accompanying Vignette for an example of how to use and interpret SummarizeGrowth. https://CRAN.R-project.org/package=growthcurver/vignettes/Growthcurver-vignette.html.
See also gcfit.
# We can check that the parameters that are found are the same # as we use to generate fake experimental data. To do so, let's first # generate the "experimental" data using the logistic equation, # e.g., absorbance readings from a single well in a plate reader over time. k_in <- 0.5 # the initial carrying capacity n0_in <- 1e-5 # the initial absorbance reading r_in <- 1.2 # the initial growth rate N <- 50 # the number of "measurements" collected during the growth # curve experiment data_t <- 0:N * 24 / N # the times the measurements were made (in hours) data_n <- NAtT(k = k_in, n0 = n0_in, r = r_in, t = data_t) # the measurements # Now summarize the "experimental" growth data that we just generated gc <- SummarizeGrowth(data_t, data_n) # Get the possible metrics for fitness proxies gc$vals$r # growth rate is a common choice for fitness gc$vals$t_gen # doubling time, or generation time, is also common gc$vals$k gc$vals$n0 gc$vals$auc_l gc$vals$auc_e gc$vals$t_mid # Compare the data with the fit visually by plotting it plot(gc)# We can check that the parameters that are found are the same # as we use to generate fake experimental data. To do so, let's first # generate the "experimental" data using the logistic equation, # e.g., absorbance readings from a single well in a plate reader over time. k_in <- 0.5 # the initial carrying capacity n0_in <- 1e-5 # the initial absorbance reading r_in <- 1.2 # the initial growth rate N <- 50 # the number of "measurements" collected during the growth # curve experiment data_t <- 0:N * 24 / N # the times the measurements were made (in hours) data_n <- NAtT(k = k_in, n0 = n0_in, r = r_in, t = data_t) # the measurements # Now summarize the "experimental" growth data that we just generated gc <- SummarizeGrowth(data_t, data_n) # Get the possible metrics for fitness proxies gc$vals$r # growth rate is a common choice for fitness gc$vals$t_gen # doubling time, or generation time, is also common gc$vals$k gc$vals$n0 gc$vals$auc_l gc$vals$auc_e gc$vals$t_mid # Compare the data with the fit visually by plotting it plot(gc)
This function finds the parameters that describe the input data's growth for a plate of growth curves. It does so by fitting the logistic curve to your growth curve measurements.
SummarizeGrowthByPlate( plate, t_trim = 0, bg_correct = "min", plot_fit = FALSE, plot_file = "growthcurver.pdf" )SummarizeGrowthByPlate( plate, t_trim = 0, bg_correct = "min", plot_fit = FALSE, plot_file = "growthcurver.pdf" )
plate |
A data.table with at least two columns. One column contains timepoints that measurements were taken (e.g., hours) and must be named "time". An optional column can be included called "blank" that contains the blank readings for background correction (make sure to select the "blank" bg_correct option if you provide a blank column). Each remaining column contains the absorbance readings from a single well in a plate. |
t_trim |
Measurements taken after this time should not be included in fitting the curve. If stationary phase is variable, this may give you a better fit. A value of 0 means no trimming. Defaults to no trimming (0). |
bg_correct |
The background correction method to use. No background correction is performed for "none". Specifying "min" subtracts the smallest value in a column from all the rows in that column, and specifying "blank" subtracts the values from the blank vector from the data_n vector. |
plot_fit |
TRUE if you want to generate a pdf file that plots all columns provided in the plate along with the growthcurver's fit. The default value is FALSE, which generates no plots. |
plot_file |
The name of the file to save the plots to if you set plot_fit to TRUE. The default file is called "growthcurver.pdf". |
The logistic curve equation is
where is the number
of cells (or the absorbance reading) at time t, is the initial
cell count (or absorbance reading), K is the carrying capacity, and r is the
growth rate.
The fitness proxies returned are the parameters of the logistic equation
and the area under the curve (a measure that integrates the effects
of , K, and r). See gcfit for more documentation on these.
This method expects that your data adhere to a particular format.
The data are provided in a data.frame
One column in the data.frame is named "time" and contains the time measurements (e.g., hours).
Each remaining column contains the readings from a single well in a plate reader. The name of the column will be used to identify the sample in the output data.
There are no missing values or non-numeric data in the data.frame.
A data.table containing the summary metrics and residual error from the fit of the logistic curve to the data. The names of the input columns are used to identify each well (or sample).
See the accompanying Vignette for an example of how to use and interpret SummarizeGrowthByPlate. https://CRAN.R-project.org/package=growthcurver/vignettes/Growthcurver-vignette.html
#Get the summary metrics for the entire plate of sample data provided #with the Growthcurver package #First, load the example data provided with Growthcurver. Note that there is #a column named "time" -- this is necessary for Growthcurver to know which #column contains the time measurements. In this dataset, the repeated #measurements from a single well in a plate are given in a column of data. myPlate <- growthdata names(myPlate) #Next, do the analysis for all the columns. summary_plate <- SummarizeGrowthByPlate(plate = myPlate) #The output is a data frame that contains the information on the best #fit for each column of data. head(summary_plate) # Use head to display just the first few rows#Get the summary metrics for the entire plate of sample data provided #with the Growthcurver package #First, load the example data provided with Growthcurver. Note that there is #a column named "time" -- this is necessary for Growthcurver to know which #column contains the time measurements. In this dataset, the repeated #measurements from a single well in a plate are given in a column of data. myPlate <- growthdata names(myPlate) #Next, do the analysis for all the columns. summary_plate <- SummarizeGrowthByPlate(plate = myPlate) #The output is a data frame that contains the information on the best #fit for each column of data. head(summary_plate) # Use head to display just the first few rows