Bioassays statistical analysis

In recent decades, the development of chemical, biochemical, and biological techniques has allowed the creation of analytical tools which can be used to facilitate the identification of the mechanisms involved in neoplastic transformation. Animal models remain, however, the most widely used approach of investigation. Cancer bioassays are usually conducted in rodents (rats and mice) and the experimental protocol takes 18-24 months and it is followed by extensive histopathological and statistical analysis. The procedure is time and... 

The sediment from Amerikahaven (site 10) was found to contain unexpectedly low contaminant levels during sampling in 1996 (see also De Boer et al., 2001). This was attributed to repeated dredging activity. The sediment was therefore sampled a second time in September 1997 at a non-dredged site. Analysis of this sediment showed considerably higher contaminant levels. These results are considered more representative of this location and were therefore used instead of the 1996 data in the multivariate statistical analysis of biomarker data. Sediment bioassays were however conducted with the material collected in 1996 and these data for location no. 10 were used for multivariate analysis when sediment chemistry was included. 

Hence, hypothesis testing (ANOVA analysis followed by multiple comparison analysis) was used to determine NOEC and LOEC values expressed as % v/v of effluent. In order to satisfy statistical analysis requirements enabling NOEC and LOEC determinations, some bioassay protocols were adjusted to make sure that there were at least three replicates per effluent concentration and at least five effluent concentrations tested. TC % effluent values were then determined as follows ... 

In the Toyama Bay Japanese effluent study, 20% endpoint effect values (e.g., LC20s for the D. magna assay and IC20s for the S. capricomutum assay), which are close approximations of TC values determined from NOEC and LOEC data (as in the Canadian study), were transformed into TU values and integrated into the PEEP formula. In applying the PEEP index concept to a designated series of wastewaters discharging to a common aquatic environment, it is paramount, of course, to use the same battery of bioassays and to report all of their toxicity responses with the same measurement endpoint and statistical analysis system (i.e., TC values for all effluents... 

Bedaux JJM, Kooijman SALM. 1994. Statistical analysis of bioassays, based on hazard modeling. Environ Ecol Stat 1 303-314. 

Another weak point of many allelopathic research studies reported in the literature is the lack of proper statistical analysis. Probably the two most critical errors are the lack of proper controls and insufficient replication. Appropriate controls need to be included even when a minimal amount of a solvent [e.g., dimethyl sulfoxide (DMSO), acetone, etc.] is used to solublilize a compound or for extraction of the test plant material. In this situation, it is also helpful to include positive controls (known compounds at similar concentrations) for comparison to the unknown or suspected allelochemicals. As discussed earlier, it is also useful to include various species so that a range of sensitivity to the test material can be observed. For example, a bioassay using seed germination might include lettuce seed, generally a sensitive species, and other species which might vary in response to the allelochemical(s) or extract. Such selection can demonstrate plant... 

Dunson DB, Haseman JK, van Birgelen APJM, Stasiewicz S, Tennant RW (2000) Statistical analysis of skin tumor data from Tg.AC mouse bioassays. Toxicol Sci 55 293-302... 

The statistical analysis of bioassay results is sometimes controversial. Problems arise when early deaths occur and the distribution of such early deaths is not uniform among study groups. This is because in groups where more animals die early, the number of animals remaining at risk for the development of tumors over the remainder of the study is reduced. Thus, the probability of finding a positive result is artificially reduced for those groups. To overcome this problem, a method... 

Statistical analysis. Results of the laboratory and field bioassays were statistically evaluated by chi-square with an acceptable significance level of at least p < 0.05i... 

The method of statistical analysis in many bioassays focuses on analyzing the number and pattern of choices made by subjects. In general, these assays will not involve truly continuous variables, but will involve counts, e.g., the number of times that each branch of an olfactometer was chosen, the number of times that upwind flight was observed, the number of eggs deposited on test or control substrates, or the number of times that test or control feeding substrates were selected. Such data often are distributed following a Poisson distribution and can... 

The method of statistical analysis of an oviposition bioassay depends on the specific bioassay and the parameters measured. In general, these assays will involve counts or frequencies (e.g., the number of eggs produced, or the number of ovipositions attempted), and analysis by goodness-of-fit G-tests will be most appropriate. General guidelines for the analysis of these data are presented in section 5.1.3. 

Hajian, G. (1983). Statistical issues in the design and analysis of carcinogenicity bioassays. Toxicol. Pathol. 11 83-89. 

Hoeven, N. van der, Kater, B.J. and Pieters, J.F. (2002). Statistical tests and power analysis for three in vivo bioassays to determine the quality of marine sediments. Environmetrics 13, 281-293. 

The q2 value of a CoMFA model, together with other statistical information from the pis analysis, provides information on the predictive capability of the model. In this study we have generated CoMFA models that describe the pharmacophore either with or without the involvement of hemin, both of which provide good q2 values. Selection of the model that most accurately depicts reality is not trivial since many variables are inherent in the cell-culture bioassay results. However, it may be... 

Once the toxicity parameters were computed to a spreadsheet yielding a table of 30 rows (effluents) and 9 columns (bioassays), we ran a principal component analysis (PCA) to check the diversity patterns of effluents and the correlation between tests. The PCA calculations were carried out using the ADE 3.6 statistical package on a Macintosh computer. ADE was developed by the University of Lyon II and by the French National Centre of Scientific Research (CNRS) common biometry laboratory. The new version ADE version 4 running on Mac and PC computers is now available on this university s internet site at http //pbil.univ-lvon 1. fr/ADE-4/... 

An in vitro bioassay can be designed in several ways, but requires statistical validity. A one point assay is not valid. The bioassay should be designed to consider factors that introduce variability, and the analysis should test such variability. A measurement series of a test sample should be compared to an equivalent series of the reference material, carefully considering the comparisons between the linear portions of the dose-response curves (Mire-Sluis et al., 1996). To test validity of a bioassay inter- and intra-assay variability should be considered in both preparation, and in the case of multiwell plates, the variability between each plate. To reduce the positional effect in plate tests, it is advisable to distribute the points on the curves randomly and also to include a reference standard in each plate (Gaines-Das and Meager, 1995). One of the most widely used techniques to validate a bioassay s performance is to include internal duplicates. The data arising from the comparison can be important in assessing the test s variability. 

Dillard RF. Statistical approaches to specification setting with application to bioassay. In Brown W, Mire-Sluis AR, eds. The Design and Analysis of Potency Assays for Biotechnology Products. Dev Biol. Basel Karger, 2002 107 117-127. 

Statistical power affects Ihe likelihood that a statistically significant result could reasonably be expected. This is especially important in studies or dose groups with small sample sizes or low dose rates. Consideration of the statistical power is often essential for reconciling positive and negative results from different studies. The impact of multiple comparisons should also be taken into account. Based on analysis of typical bioassays in which both sexes of two species were included, studies in which there is only one significant result that falls short of the 1% level for a common tnmor shonld be treated with cantion (EPA 2005a Hasetnan 1983). 

The statistical analyses of the data concerning the four caterpillar bioassays involves an analysis of variance (ANOVA) for a completely randomized design (CRD) followed by Dunnett s test for each experiment. The ANOVA are done to get a preliminary feel if the treatments had any effect on the test organism. Dunnett s method is chosen because we are interested in determining whether the mean of the control group is significantly different than each of the means of the treatments. Experimental treatments can consist of crude extracts or pure compounds. 

---