Data Analysis and Interpretation

BG Medicine’s data analysis and interpretation processes are seamlessly integrated with its bioanalytical sample analysis platforms to form a cohesive, robust workflow. Statistically suitable experimental designs are developed and tailored to address the objectives of each project. Customized data processing, data normalization and data alignment algorithms allow BG Medicine to integrate disparate metabolomic and proteomic datasets. We recognize that the challenge in molecular biomarker studies lies not in the large volumes of data but in their heterogeneity, and we integrate molecular as well as clinical, imaging, and other data sources in a coherent analysis.


Statistical Data Analysis

BG Medicine’s biostatistics staff is fluent in the new mathematical language of datasets characterized by extremely large numbers of diverse analytes measured in a single experiment, together with appropriate clinical outcome measures and study covariates. For each discovery project, statistical modeling and analyses are tailored and implemented in a secure and audit-compliant database environment, yielding a highly efficient, rapid and scalable data analysis pipeline.


Biological Data Interpretation

BG Medicine’s bioinformatics staff specializes in integrating multiple sources of biological knowledge to provide expert interpretation and elucidate and contextualize molecular measurements. Our staff leverages a variety of proprietary tools to analyze and interpret study data, such as our own BTE™ databases of biochemical pathways and our Seer™ data interface (screenshot below) for traversals of integrated datasets and individual biochemical reactions.

In the analysis of molecular and clinical data BG Medicine leverages the inherent biological variability among study subjects to detect subtle and early modulations of biochemical pathways and mechanisms that underlie the observed molecular changes. This requires extremely stringent tolerances on bioanalytical platforms, namely that the measurement variances be consistently less than individual-to-individual biological variability. This approach, as exemplified by our Correlation Network™ framework analysis, is key to identifying biomarkers that are specific to the biological phenomena and mechanisms of interest.

IT Infrastructure

BG Medicine has developed a custom, proprietary IT infrastructure for its biomarker discovery and validation programs. Ranging from laboratory processes to data analysis and statistics to bioinformatics, our IT infrastructure provides an integrated, flexible and powerful data management and software environment. BG Medicine’s IT systems, built on Oracle® backbone technology, are routinely audited for compliance with relevant regulatory standards. Secure data access as well as multiple redundancies and secure off-site co-location data centers ensure the integrity and safety of all data and related captured processes.