Animal Models

Question 1

Definition of animal model

Answer 1

• a model in which normative biology or behavior can be studied, or in which a spontaneous or induced pathological process can be investigated, and in which the phenomenon resembles in at least one respect the same phenomenon in human or other species of animal

Question 2

Groups of animal models for the study of human disease:

Answer 2

• induced (including transgenic)
• spontaneous (naturally occurring genetic variants, inborn, non-experimentally induced)
• negative (occasions when a certain disease does not develop, or animals are insensitive to a certain stimulus)
• orphan (a disease is identified and studied in animal models with the knowledge that a human counterpart could be identified at a later stage)

Question 3

Three different philosophies of knowledge acquisition:

Answer 3

Inductive-hypothetico-deductive philosophy
- observation of all possible facts in order to gather relevant data from which a hypothesis is developed inductively (in conflict with Hume’s view of the untenability of inductive reasoning, i.e. extracting a general law form repeated observations)
The deductive philosophy of Karl Popper
- no theory can be shown to be true, but some can be shown to be false
- emphasis on the falsifiability of scientific theories
- deduction: a working hypothesis is arrived at by choosing among competing hypotheses, many of which can be disregarded as they cannot be shown to be false
The paradigm shift philosophy of Thomas Kuhn
- science progress by sudden change followed by a period of consolidation

Question 4

What is an experimental unit?

Answer 4

• it’s defined as the unit which can be assigned independently to one of the treatments. This is the unit which will be subjected to statistical analysis.

Question 5

Estimates needed to calculate sample size:

Answer 5

• required significance level
• required statistical power (1-probability of type II error)
• minimum significant treatment effect of interest
• anticipated variation

Question 6

Mead’s “resource-equation method” for calculating sample size:

Answer 6

N-1 = T + B + E

• N - number of experimental units
• T - treatment degrees of freedom (treatments - 1)
• B - # of block + covariates - 1
• E - error; N - (T+B)

Question 7

Mead’s recommendations of df for animal experiments

Answer 7

• fewer than 10 treatment combinations is considered wasteful
• more than 50 - too complicated

Question 8

Inter-individual variation or “error” is made up of two components:

Answer 8

• • variation due to analytical errors, and “time” effects at the time that the measurements are made
• intrinsic individual variability

Question 9

Ways of reducing intrinsic variability:

Answer 9

• use of isogenic strains (genetically uniform animals)
• use of SPF animals
• environmental & nutritional control
• selection for uniformity (seriously?! selection of animals based on observed characteristics)
• acclimatization

Question 10

Reduction of error by stratification or blocking

Answer 10

• putting similar animals into blocks and randomly assigning those to treatment conditions