Statistical thinking in antibiofilm research Cord Hamilton Al Parker Marty Hamilton MBL and SBML: 23 October
Topics (presenter) Calculating LR and the within-experiment standard error of LR (Cord) Using data from repeated experiments to find more reliable LR values in the future (Al) Analysis of dilution series counts (Marty) 2
Log Reduction (LR) for a Quantitative Assay V c = viable cell density of biofilm grown in the absence of antimicrobial treatment V d = viable cell density of biofilm grown in the presence of the disinfectant Log Reduction = log 10 (V c ) - log 10 (V d ) 3
Numerical Example V c = 10 7 & V d = 10 Log Reduction = log 10 (10 7 ) - log 10 (10) LR = LR = 6 Interpretation: disinfectant killed % of the bacteria 4
Calculating LR when there are multiple coupons = mean of control log 10 densities = mean of disinfected log 10 densities Log Reduction = C D C - D 5
Example: Mean of logs for 3 disinfected coupons Coupon Density log 10 Density (i) cfu / cm 2 (D i ) 1 9.6· · · Mean = = mean density = 4.09∙10 4 log of mean density = 4.61 D 6
Example: Control coupons Coupon log 10 Density (i) (C i ) C =
Calculating LR when there are multiple coupons = & = Log Reduction = C D C - D = LR =
Within-experiment standard error (SE) of the LR S c = variance of control log 10 densities S d = variance of disinfected log 10 densities n c = number of control coupons n d = number of disinfected coupons SE of LR = (within-experiment) S ncnc S ndnd c 2 d 2 2 2 9
Example: Calculating SE for single reactor experiment S c = and n c = 3 S d = and n d = 3 SE = 2 2 =
Uncertainty in LR Estimate LR ± SE= ± or 3.06 ± 0.39 or 3.1 ±
Log Reduction ± SE Experiment 12 RDR biofilm : 5 ppm chlorine for 10 minutes
Experiment repeated 3 times, each using three control and 3 disinfected coupons 13
Statistical summary for data from 3 experiments, with 3 control and 3 disinfected coupons per experiment log densitymean log densitySD log densityStandard error of Expcontroldisinfectedcontroldisinfected log reductioncontroldisinfectedlog reduction Pooled within-experiment SD of the control log density: Pooled within-experiment SD of the disinfected log density: Between-experiment SD of the log reduction:
S n c m c 2 + Formula for the SE of the mean LR, averaged over experiments S c = within-experiment variance of control coupon LD S d = within-experiment variance of disinfected coupon LD S E = between-experiments variance of LR n c = number of control coupons n d = number of disinfected coupons m = number of experiments S n d m d 2 + S m E 2 SE of mean LR = 15
Formula for the SE of the mean LR, using estimated standard deviations n c m n d m m 2 SE of mean LR = 2 Pooled within-experiment SD of the control log density: Pooled within-experiment SD of the disinfected log density: Between-experiment SD of the log reduction:
Choosing the numbers of coupons and the number of experiments. Table cell is the the SE of the mean LR. Shaded SE values are designs requiring 24 coupons. no. control coupons (n c ):23612 no. disinfected coupons (n d ):23612 no. experiments (m)
Dilution series and drop plate technique Source: BiofilmsOnline Counted dilution 32 colonies 10 18
Find the fraction of initial beaker volume in each of the dilution tubes Source: BiofilmsOnline Beaker: contained all cells from coupon fraction of beaker volume in tube 10 19
Estimated number of cells in beaker = cfu count divided by the volume fraction plated Beaker: contains all cells from coupon fraction in tube Plated 50 μl from tube; plate contains a fraction 50/10000 = 5 x of the volume in the tube. f = (5 x ) = 5 x Estimate: 32/(5 x ) = 6.4 x
Dilution series and filter technique: pooling data from two tubes 9 ml filtered 10 ml filtered Count 20 fields on each filter; corresponds to 0.02 of filter area f = x 0.9 x 0.02 = 1.8 x f = x 1.0 x 0.02 = 2.0 x cfu 39 cfu The 460 cfu corresponds to this fraction of the beaker volume: f = 1.8x x10 -6 = 2.0 x Estimate for beaker = 460/(2.0x10 -5 ) = 2.3 x