Research and Application: Examples - Biomedical Informatics in Translational Research - page 224

gies that often co-occur with DCIS in general [46]. Research by other groups has
noted that allelic loss increases with histological grade [47]. Furthermore, the ability
to discriminate high-grade from low-grade DCIS by genomic characterization, and
the implication of separate genomic pathways for these two lesions, has previously
been reported [48]. However, the expanded clinical and epidemiological data avail-
able in the WRI data repository will enable researchers at WRI to expand on the
results detailed in this chapter. The challenge facing WRI researchers is taking these
results to the next level: the identification of DCIS characteristics (genomic,
proteomic, and tissue microenvironment) and patient characteristics (epidemiologi-
cal, comorbidities, reproductive history) that predict likely recurrence or

progression to invasive breast cancer.

10.4 Conclusions

Both of the case studies presented in this chapter highlight medical and
multidisciplinary scientific contributions necessary to enable translational research
into human disease. Because of the multiple sources of data (clinical, “omic,” epide-
miological, and so on) and typically large quantity of data used in translational
research, the new discipline of biomedical informatics is arising to facilitate this type
of research. In this chapter we detailed how clinical data collection and storage,
genomic analyses, novel algorithm development, and most importantly access to
large repositories of human data for analysis enabled both deCODE and WRI to
make novel discoveries in human disease.
In this chapter we also have demonstrated how using data from multiple sources
enables the discovery of previously unknown molecular correlates to human disease.
In the case of deCODE we saw how identification of previously unknown genes led
to the discovery of new drugs and new diagnostics that are making their way
through clinical trials into clinical use. Without access to both human sample data
(DNA), clinical data, and genealogic information, and the ability to integrate these
three data sources, deCODE would not have been able to discover multiple genetic
associations to complex human diseases. The data and tissue resource at WRI has
enabled genomic analysis and data-mining studies that show an emerging picture of
the genomic and tissue environment factors that can differentiate between
high-grade and low-grade cancer. WRI now faces the challenge of translating this
knowledge into a clinically useful tool.
Where this chapter looked at how biomedical informatics enabled research that
could be translated into clinical use, the next chapter will look at how the clinic can
motivate translational research projects.

References

[1] http://www.decode.com.

[2] Gulcher, J., and Stefansson, K., “Population Genomics: Laying the Groundwork for Genetic

Disease Modeling and Targeting,” Clin. Chem. Lab. Med. , Vol. 36, 1998, pp. 523-527.