Building biostatistics

As drug companies incorporate pharmacogenetics into their development mix, they are finding themselves short of key skills. The genome-wide analysis at the heart of pharmacogenetics needs scientists comfortable at the intersection of statistics, mathematics, informatics, genetics and biology. They also need to develop and apply large-scale analysis, and ensure that the results are meaningful and relevant.

Essential techniques in this field include text mining, systems biology, advanced multivariate statistics of gene–gene interaction, genetic epidemiology and statistical genetics, statistical data mining for microarray analysis, genome mining for gene discovery, and mathematical modelling of complex biological systems.

This set of skills is difficult to find in a single person, and interdisciplinary training programmes are only beginning to emerge. Also, in the past, statistics and applied mathematics have become separated; biology needs both.

Even this demanding set of technical skills is not enough. The ideal pharmacogenetics researcher needs excellent communication skills to talk to physicians about phenotypes, to discuss algorithms with mathematicians and explore data analysis and integration with biostatisticians and bioinformaticians.

As academia hasn't been producing these kinds of people, industry may have to make its own. That is the strategy the Serono Genetics Institute (SGI) in Evry, France, is taking — but with a twist.

The SGI needs to boost its 100-scientist staff by 10% this year, particularly in biostatistics for complex-trait genetics and in bioinformatics, to support its pharmacogenetics research.

How to accomplish this? First, the SGI will use as many recruiting tools as possible to find senior and experienced scientists. It is also looking for young scientists who excel in a skill the company needs and have set up their research or career paths early, or shown curiosity about interdisciplinary research. And to meet the target of ten more experts in biostatistics and bioinformatics this year — and probably more in the future — the SGI will train its own future scientists by offering PhDs and postdocs the possibility of tenure.

So, young scientists should have the courage to build up an unconventional and interdisciplinary career that could lead them into pharmacogenetics.