Name
Title | ||
|---|---|---|
Data Processing Effects on the Interpretation of Microarray Gene Expression Experiments |
Abstract | ||
|---|---|---|
Motivatio n :M icroarray gene expression dat ai sc ollected at an increasing pace and nu- merous methods and tool se xist fo ra nalyzing this kind of data. Th ea im of this study is to evaluate th ee ffect of th eb asic statistical processin gs teps of microarra yd at ao n th efi nal outcome fo rg ene expression analysis; these effect sa re most problematic fo r one-channel cDN Am easurements, bu ta ls oa ffect other types of microarrays, espe- cially when dealin gw ith grouped samples. It is crucia lt od etermine an appropriate combinatio no fi ndividual processing steps fo ra give nd ataset in orde rt oi mprove th e validity an dr eliability of expressio nd at aa nalysis. Results :W ea nalyzed al arge gene expression dat as et obtained fro mao ne-channel cDN Am icroarra ye xperiment conducted on 83 human samples that have been clas- sified into four Osteoarthritis related groups. We compared different normalization methods regarding th ee ffect on th ei dentificatio no fd ifferentially expressed genes. Furthermore ,w ec ompared different methods fo rc ombining spot p-values into gene p-values, an dp ropose Stouffer' sm ethod fo rt hi sp urpose. We de v eloped severa lq ual - ity and robustness measures which allo wt oe stimat et he amount of errors made in th e statistical dat ap reparation. Conclusion :T he apparently straight forwar ds teps of gene expression dat aa nalysis, i.e .n ormalizatio na nd identificatio no fd ifferentially expressed genes, can be accom- plishe db yn umerous different methods. We analyzed multiple combinations of an um- ber of methods to demonstrate th ep ossibl ee ffect sa nd therefor et he importance of th e singl ed ecisions take nd urin gd at ap rocessing. An overvie w of these effect si se ssentia l fo rt he biological interpretatio no fg ene expression measurements. We give guidelines and tool sf or evaluatin gm ethods fo rn ormalization, spot combinatio na nd detectio no f differentially regulated genes. |
Year | Venue | Keywords |
|---|---|---|
2005 | German Conference on Bioinformatics | gene expression,data processing |
Field | DocType | Citations |
Complementary DNA,Data processing,Normalization (statistics),Gene,Biology,Gene expression,Microarray analysis techniques,Bioinformatics,DNA microarray,Gene expression profiling | Conference | 1 |
PageRank | References | Authors |
0.49 | 14 | 4 |
Name | Order | Citations | PageRank |
|---|---|---|---|
| Katrin Fundel | 1 | 492 | 23.60 |
| Robert Küffner | 2 | 512 | 59.39 |
| Thomas Aigner | 3 | 1 | 0.49 |
| Ralf Zimmer | 4 | 1 | 0.49 |