Systems Biology Data
Preprocessing

A Java-based data management tool, Xmanager, was developed for importing, preprocessing, and managing multi-conditional experimental next generation sequencing, microarray, protein abundance from MPSS and other data. XManager semi-automatically imports the time-course experimental gene array, protein and physiological experimental data into our relational database. All entries will be annotated according to a unifying symbol and ID system during the process. Xmanager is suitable for many data formats, including GenePix, Affymetrix (GCOS), Codelink, Agilent or proteomics data from mass spectrometry or next generation sequencing and is easily extendable to include any other measurement format/platform.

Xmanager is capable of preprocessing the experimental data such that related data obtained from different experiments or platforms can be compared and modeled as a whole. The preprocessing engages a series of steps including normalization (global scaling, mean-, median-, Lowess-normalization), quality assessment and filtering (coefficient of variance, signal-to-noise ratio, background randomness, signal correlation), channel/array merging and classical statistic analyses (z-test, t-test, Bayesian statistics, fold change, boxplot etc.). Careful preprocessing is an important step before proceeding into the analysis and modeling steps. Xmanager will then deliver normalized, annotated data sets into the database for use by other analyses and modeling steps. Process metadata will also be preserved so that all pre-processing steps are recorded for future reference and recreation if needed.