NOEC {mixtox}

NOEC and LOEC Calculation

Description

calculating the NOEC and LOEC using Dunnett's test

Usage

NOEC(x, rspn, blankC = FALSE, sigLev = 0.05, alternertive = 'B', sav = FALSE)

Arguments

x

a numeric vector of experimental concentrations

rspn

a numeric matrix of experimental responses with at least three replicates.

blankC

TRUE if rspn contains responses of blank control. The default is FALSE.

sigLev

the significance level for Dunnett's test. The default is 0.05.

alternertive

the alternative hypothesis: "U"=upper one-sided test; "B"=two-sided test(default).

sav

TRUE: save output to a default file; FALSE: output will not be saved; a custom file directory: save output to the custom file directory.

Details

Dunnett's test (Dunnett, 1964) is performed to compare the treatment groups with the blank controls. The critical constants (store in DTcv) were calculated using step-down Dunnett test procedure. Three significance level (0.01, 0.05, and 0.1) are supported. ## Q: One dataset has four blank controls (C1, C2, C3, C4) and one treatment has three replicates (T1, T2, T3), ## another treatment has five replicates (R1, R2, R3, R4, R5), how to arrange the response matrix (rspn)? ## A: Label the missing values as NA, the response matrix (rspn) can be arranged as follows:

C1 C2 C3 C4 NA
T1 T2 T3 NA NA
R1 R2 R3 R4 R5

The adjustation of critical value for the unequal variances or unequal number of control and replicates is skipped in this program.

Value

mat

information on Dunnett's test.DT: Dunnett's test values; DTcv: critical values for Dunnett's test at the significance level of sigLev.

noec

non-observed effect concentration (NOEC).

loec

least-observed effect concentration (LOEC).

sigLev

the significance level used in the Dunnett's test.

DF

the number of treatments and degree of fredom.

sav

TRUE: save output to a default file; FALSE: output will not be saved; a custom file directory: save output to the custom file directory.

Note

x a vector of concentrations or levels in an ascending order. response matrix with at least 3 replicates. if the response matrix (rspn) contains blank controls (blankC = TRUE), the blank controls should be allocated in the first tow of rspn matrix. missing values should be labled as NA.

References

Dunnett, C.W., 1964. New tables for multiple comparisons with a control. Biometrics 30, 482-491.

Examples

## example 1
# calcualte the NOEC and LOEC of heavy metal Ni(2+) on the MCF-7 cells at the default significance 
# level of 0.05
x <- cytotox$Ni$x
rspn <- cytotox$Ni$y
NOEC(x, rspn)

## example 2
# calcualte the NOEC and LOEC of Neomycin sulfate on the phtotobacteria at the significance 
# level of 0.01
x <- antibiotox$NEO$x
rspn <- antibiotox$NEO$y
NOEC(x, rspn, sigLev = 0.01)