Comparative Research is Not the Same as Predictive Research

| by Dr Ray Greek

In the June 2012 issue of Nature Medicine, an article was published in the news section touting efforts by veterinarians to work with physicians. The purpose in the cooperative effort is to gain more understanding of disease, specifically diseases that affect humans. I wrote the following letter to Nature Medicine, which was declined for publication.

Gammon (1) discussed the utility of veterinarians and physicians working together in order to gain further understanding of diseases that affect humans and nonhuman animals. While this sounds promising in principle, the reality is that animals have been used as models of human disease for centuries. Animals that share diseases characteristics with humans are referred to as spontaneous models (2) and are well described in the literature (3). Unfortunately, the results for predicting human response are unimpressive (3). Animal models in general consistently yield a low positive and negative predictive value (4). Face validity of a model does not imply construct validity and while face validity is commonly seen, construct validity is not. For example, the genes that result in Sanfilippo syndrome and phenylketonuria in humans do not cause these condition in macaque monkeys (5). Although the discrepancy between construct validity and face validity for animal models has long been appreciated, complexity science and developments in evolutionary biology such as evolutionary and developmental biology and insights from the Human Genome Project have allowed a fuller appreciation for why this is the case.

Popular Video

Here is the footage of Snoop Dogg shooting President Trump in his new music video. Did he cross a line?

Humans and animals are complex evolved living systems and thus exhibit properties such as emergence, robustness, redundancy, nonlinearity, modularity, dependence upon initial conditions, hierarchal levels of organization, and the whole is greater than the sum of the parts. Thus, there are limits to what can be learned about a complex system using reductionism and thus limits on what can be extrapolated between complex systems (6). Evolution has affected all the properties of a complex system (7). For example, gene pleiotropy, alternative splicing, and gene redundancy have implications for ascertaining gene function for another species. Convergent evolution has resulted in the same trait being developed by different pathways, for example the eye of humans and octopuses (8). Differences in gene network, gene regulation and expression, the presence of modifier genes or mutations, and environmental influences can result in very different outcomes from the same set of genes. This concept is perhaps illustrated best by the fact that monozygotic twins differ in disease susceptibility, in part secondary to very small differences between the genomes (9). Relying on animal models to predict human response to drugs and disease while simultaneously pursuing personalized medicine because of intra-species differences suggests an inconsistent approach to research for human disease.

Animals can be successfully employed in many aspects of science and research. For example, there is no argument regarding the value of using of animals in comparative research and animals are routinely used to supply tissues and other products for humans. Were the point of the Gammon article that veterinarians and physicians should collaborate more frequently in comparative research, such would be noncontroversial. Implying that animal models, spontaneous or otherwise, can be predictive for human response to drugs and diseases like cancer however, is not supported by current science (10).

Popular Video

Here is the footage of Snoop Dogg shooting President Trump in his new music video. Did he cross a line?


1.         Gammon, K. Nat Med 18, 847-849 (2012).

2.         Hau, J. Animal Models. in Handbook of Laboratory Animal Science. Second Edition. Animal Models, Vol. II (eds. Hau, J. & van Hoosier Jr, G.K.) 1-9 (CRC Press, Boca Rotan, 2003).

3.         Tran, P., et al. Mol Genet Metab 76, 297-304 (2002).

4.         Shanks, N., Greek, R. & Greek, J. Philos Ethics Humanit Med 4, 2 (2009).

5.         Gibbs, R.A., et al. Science 316, 222-234 (2007).

6.         Greek, R. & Shanks, N. Stud Hist Philos Biol Biomed Sci 42, 542-544 (2011).

7.         Greek, R., Menache, A. & Rice, M.J. Personalized Medicine 9, 47-64 (2012).

8.         Kirschner, M.W. & Gerhart, J.C. The Plausibility of Life, (Yale University Press, 2006).

9.         Maiti et al. PLoS ONE 6, e17125 (2011).

10.       Wall, R.J. & Shani, M. Theriogenology 69, 2-9 (2008).