Challenges and Opportunities in Clinical Bioinformatics

Abstract

Clinically applied bioinformatics faces many specific challenges. Many of these are related to challenges faced by discovery bioinformatics, which is confronted by similar issues of data complexity, data scale, and the lack of statistical power to address key problems. These include issues of the scale and variety of data being handled and annotated, and the associated proliferation of errors of data annotation, and errors of statistical inference. Modelling issues include the choice of methods, and the flaws associated with overly simplistic and overly complex approaches.\rData aggregation among researchers and clinicians and patients is likely to represent a key step forward, but the main clinical gains are likely to emerge from the aggregation of relatively homogeneous data types, associated with clear prior hypotheses. It may be less useful to integrate analytical and predictive approaches across many different complex data types in smaller groups of subjects.\rClinical bioinformatics needs to be integrated into new regulatory paradigms for incorporation of knowledge into healthcare. We need to explore possibilities of one-person trials integrated with genotypic data but the theoretical and practical frameworks for such approaches are not worked out.