| |
3DNA Array 900, Array 350 and Array 50
(Version 2) Kit Troubleshooting Guide |
|
| |
Section V. Data Problems
|
| Symptom |
Cause |
Resolution |
| 1. Strong signal/poor differential. |
A. cDNA hybridization and/or wash temperatures too low. |
The hybridization stringency is defined by the cDNA hybridization and wash conditions. The signal and differential should be adjusted at this step. High signals are not necessarily correct signals. If the hybridization or wash temperatures of the cDNA hybridization are not stringent enough, signal may result from non-specific hybridization of synthesized cDNA to the wrong target. This generally leads to "masking" of otherwise highly expressed genes. If this occurs, increase the hybridization temperature by 2-5°C and/or increase the wash temperature by 5-10°C. The hybridization and wash temperatures outlined in the protocol are an approximation of the optimal temperatures for oligo and cDNA arrays. The end user should use this information as a starting point for optimization of the hybridization and wash conditions. For additional recommendations please consult the protocols or call Genisphere Technical Support. |
|
 |
| |
B. Insufficient poly dA/dT blocking. |
Genisphere's proprietary LNA dT blocker (Vial 9) employs the use of Locked Nucleic Acids, or LNA, to decrease the likelihood of hybridization of synthesized cDNA PolyT regions to Poly A sequences printed on the array (as either control or part of the spotted cDNAs). When the poly A regions of spotted cDNA are not effectively blocked, the poly dT sequences that are part of synthesized cDNA in the hybridization mixture will bind specifically to these areas. Any specific poly dA/dT hybridization will result in compressed differential at that feature. For further details, refer to the “Buffer Components” section (Array 350 and 50 V2 kits) or "Hybridization Conditions" section (Array 900 kit) of the protocols. |
|
|
| |
C. Insufficient repetitive sequence blocking. |
Repetitive sequences can result in compressed differentials. Use of a competitive nucleic acid, such as Cot-1 DNA, may resolve this problem. |
|
|
 |
|
|
| |
| Symptom |
Cause |
Resolution |
|
|
|
| 2. False Positives. |
A. Feature auto-fluorescence. |
Many spotting buffers, particularly 3XSSC, show a high level of fluorescence, mostly in the Cy3 channel. This signal can carry through the hybridization, affecting the data. Prescan the array, before hybridization, to determine any pre-existing signal. Some array processing procedures may help to remove any remaining spotting buffer from the array. Contact Genisphere Technical support for further recommendations on eliminating pre-existing feature fluorescence. |
|
|
|
B. Unpurified PCR clones. |
Unpurified PCR clones that have been spotted onto arrays may promote unusual and unpredictable hybirdizations. Make certain to use a clean-up column, such as the QIAquick PCR Purification Kit from QIAGEN, on all PCR products that will be spotted onto microarrays. |
|
|
|
C. 3DNA hybridization and/or wash temperatures too low (oligo arrays). |
As with all labeling methods, Genisphere labeling kits may present some level of false positive signal on certain array types. In particular, 3DNA molecules have shown affinity for some oligo targets under non-ideal hybridization conditions. This phenomenon can be reduced or eliminated by maintaining the appropriate 3DNA hybridization and wash conditions. For more information, refer to the "3DNA Hybridization" section of the protocols. If problems persist, please contact Genisphere's Technical Support. |
|
|
|
D. Excessive primer used in RT reaction (oligo arrays). |
Using excessive amounts of primer in the reverse transcription reaction can promote nonspecific false positives. Using the appropriate amount of primer may eliminate this problem. For further information, refer to the "cDNA Synthesis" section of the protocol. If problem persists, please contact Genisphere's Technical Support. |
|
| |
| Symptom |
Cause |
Resolution |
|
|
|
| 4. Poor reproducibility. |
A. Array inconsistancies. |
Array spotting and processing are prominent sources of reproducibility problems. Always use, in any one experiment, successive arrays from the same printing batch. Running replicate arrays is a common method of overcoming array inconsistancies. The optimal number of replicates is widely debated among statisticians in the field. |
|
|
|
B. RNA inconsistancies. |
RNA preparations, even from the same cell source, can often result in variablility in purity and quantity. These inconsistancies can result in variability in microarray data. When comparing microarray data, note the lot of RNA used in each data set. Replicates may be needed to increase the confidence of the gene expression data. The optimal number of replicates is widely debated among statisticians in the field. |
|
|
|
C. Failure to normalize data appropriately. |
Failure to normalize data in a two color experiment can have a tremendous impact on calculated gene expression changes and often yield apparent poor reproducibility. Use of one of the many normalization methods reported in the literature will facilitate the ability to effectively compare data with future data as well as that reported by other laboratories. |
|
|
|
D. Inappropriate hybridization or wash conditions. |
The hybridization and wash conditions outlined in the protocols are intended to be used as a guide for further optimizations. Use of suboptimal hybridization or wash conditions (both temperature and time) may lead to variable results. As such, optimal conditions should be imperically determined by the end user. |
|
 |
|
|
|
