A Quick Guide to Large-Scale Genomic Data Mining 英文参考文献.docVIP

Education AQuickGuidetoLarge-ScaleGenomicDataMining CurtisHuttenhower*,OliverHofmann DepartmentofBiostatistics,HarvardSchoolofPublicHealth,Boston,Massachusetts,UnitedStatesofAmerica Introduction coexpress in high-throughput data repositories?Underwhatexperimental conditions,orinwhichtissues? as great or greater a concern as data processing: plain text or XML storage formats, while conveniently human-read- able, can waste unsustainable amounts of spaceforlargerepositories. For the first several hundred years of research in cellular biology, the main bottleneck to scientific progress was data collection. Our newfound data-richness, however, has shifted this bottleneck from collection to analysis [1]. While a variety of options exists for examining any one experimental dataset, we are still discov- eringwhatnewbiologicalquestionscanbe answeredbyminingthousandsofgenomic datasets in tandem, potentially spanning differentmolecularactivities,technological platforms, and model organisms. As an analogy, consider the difference between searching one document for a keyword andexecutinganonlinesearch.Whilethe tasksareconceptuallysimilar,theyrequire vastly different underlying methodologies, Bringing large quantities of genomic data to bear on such questions involves threemaintasks:establishingmethodology for efficiently querying large data collec- tions; assembling data from appropriate repositories; and integrating information fromavarietyofexperimentaldatatypes. Since the technical [5–7] and methodo- logical[8–10]challengesinheterogeneous Solutions to these technical issues in- cludesoftware anddataaccess methodol- ogies specifically tailored to large-scale data manipulation. Three broad catego- ries of solutions exist: Web applications that aggregate information from multiple sources, programmatic APIs that allow sophisticated computational queries of data integration have been discussed individual large data sources, and do-it- yourself solutions that rely on manually