Biostatistics and Bioinformatics for Molecular Medicine
Thanks to modern assay techniques, biology is transforming into a "data-rich" science: With high-throughput sequencing, mass spectrometry, automated drug and perturbation screens and other "big data" assay techniques, we can now at the same time get a "bird's eye" overview as well as plenty of detail on a biological sample. For medicine, this means that doctors will soon be able to acquire an unprecedented richness of raw information on a patient's body and his or her disease. This will help us understand, why different patients with the same disease often react so differently to the same treatment, why some patient recover from a dangerous disease and others, though treated alike, succumb to it -- and ultimately this will pave the way to medical treatments chosen for a specific patient and not just for a disease.
The vast amounts of raw data produced by modern high-throughput assays, however, are of little use without powerful bioinformatics and biostatistics methods to process, analyse, and interpret them.
The Anders group will focus on developing the computational tools that clinical researchers need to find needles of biological and medical insight in big haystacks of data.
Computational scientists provide crucial expertise to research in molecular medicine that is complementary to the skill set that a typical biologist or physician has learned in her or his studies: We know how how to process, organise and explore big data sets, and transform data such that it can be viewed from "just the right angle" to gain new insights into biological systems (exploratory data analysis, EDA) and we know how to ensure that findings are true and reproducible rather than the result of random chance or wishful thinking (inferential statistics).
Exploring big data requires tight collaboration, and especially effective communication, between computational scientists and biologists or clinical researchers. While only the latter have the in-depth knowledge and intuition of the system under study to know which specific questions should be asked, we know how what questions can be asked from a big data set, and, importantly, we can invent new investigative strategies to ask entirely new kinds of questions that were not accessible before.
For my group, I aim to assemble an interdisciplinary team, drawing not only from bioinformatics and statistics, but also from computer science, physics, engineering and other quantitative subjects, in order to cover a broad field of expertise on methods that can be translated, adapted and expanded for use in molecular medicine.
We will strive to strike a balance between two modes of research: On the one hand, we will craft tailored analysis methods for specific advanced experiments carried out by our wet-lab and clinical colleagues in Helsinki and elsewhere. On the other hand, we will package such approaches into generally useable software tools that are modular, documented well and easy to use, to ensure that even research group with little or no biostatistics expertise can download and use them, thus making cutting-edge biostatistical methodology available to the research community at large.
The following exemplary project outlines are meant to give a flavour of the research I want to tackle once my group gets started.