Skip to main content

Expertise

Our Mission

The mission of the DBB is to foster and facilitate excellence in UMSOM quantitative pre-clinical and clinical cancer research by providing in-depth knowledge of and access to state-of-the-art biostatistics and bioinformatics methods. Domain expertise of DBB faculty covers a variety of fields related to pre-clinical, clinical and population level biomedical research, but a particular strength is clinical trial design, biomarker discovery & validation studies and epidemiology.

The Center currently offers support with:

Study Design and Power Analysis. Feasibility study and sample size based on the state of the art methodology such as multi-stage and adaptive designs. In addition, we design and perform randomization, sensitivity and simulation studies to evaluate study design properties, and sample sizes needed to achieve study objectives, endpoints and control selection.

Statistical Analysis. The data analysis plan is provided in the protocol (before data collection takes place) and followed. Assumptions underlying the methods are assessed and the most appropriate statistical method is used for analysis. Beyond major statistical software packages such as SAS, Splus, Stata, StatXact and East, more complex statistical and bioinformatics tool are developed for specific research project (in R or Matlab or SAS).

Interim Analyses and Data and Safety Monitoring for Clinical Trials. Analyses and adaptive designs (including Bayesian designs) consistent with statistical principles for multiple testing and interim analyses.

Statistical bioinformatics. Another main focus is in statistical bioinformatics and the analysis of ‘Big Data’ including systems biology, gene regulatory network inference, module biomarker discovery, network biology, data integration, systems pharmacology, personalized medicine, disease-drug-gene association discovery, pattern recognition, image analysis and data mining.

Reports and Publications. Data analysis is typically summarized in a statistical report in a form appropriate for manuscripts.