1. GENERAL QUESTIONS/MISCELLANEOUS

• 1.1 What is BarleyBase?

• 1.2 How is  BarleyBase funded?

• 1.3 Does BarleyBase only store Affymetrix barley GeneChip data?

• 1.4 How is BarleyBase integrating expression data with other genomic data

• 1.5 I find some vocabulary not familiar, where can I find their exact meanings?

2. EXPERIMENT SUBMISSION

• 2.1 What's the submission interface for BarleyBase?

• 2.2 Is the  submission tool, BarleyExpress, consistent with the  MIAME standards?

• 2.3 What is needed before I start my submission?

• 2.4 Which operating systems can be used for making submissions?

• 2.5 Which browser works best  for the submission process, how do I set my browser to make submission?

• 2.6 Why does BarleyExpress require experiment factor and factor levels?

• 2.7 Must I finish a submission at a single time? Can I stop at some point and resume submission later?

• 2.8 Can I add more experiment factors or factor levels to my finished submission?

• 2.9 Can I change the names of the experiment factors/level?

• 2.10 Any recommendations for making a smooth submission? 

• 2.11 I have many samples, and they are very similar, how to submit them quickly? 

3. MANAGEMENT OF ACCOUNT, EXPERIMENTS AND DATA ACCESSIBILITY

• 3.1 Why register with BarleyBase? How to register with BarleyBase? ? 

• 3.2 What's the purpose of group management?  

• 3.3 Will my submission be public immediately? Can I control who will be able to access my submission before my results are published?

• 3.4 How can I grant access to my private data to other users? Can I use delete users from the access list?

• 3.5 What if I designated my data as private when making my submission, but later on I need to let selected people to access it?

• 3.6 Can I get all data for an experiment with batch download? What types of data are downloadable?

• 5.1 Can I analyze the data sets from a query using BarleyBase? If so, what analysis is available?

• 5.2 Can I select a sub-cluster from a dendrogram to do more detailed analysis?

• 5.3 How is the online data analysis carried out at BarleyBase?

• 5.4. Is data transformation and normalization important? How to decide when this is essential?

• 5.4. What types of microarray data visualization is offered in BarleyBase?

1.1 What is  BarleyBase?

BarleyBase is a MIAME-compliant, MySQL relational database, which serves as a public repository for raw and normalized expression data from the Affymetrix Barley1- and ATH1-121501-GeneChips.

1.2 How is BarleyBase funded?

BarleyBase is funded by USDA-NRI grant no. 2002-35300-12619.

1.3 Does BarleyBase store only Affymetrix Barley GeneChip data?

No. BarleyBase accepts data obtained with the Barley1 GeneChip, and from Arabidopsis, obtained from ATH1-121501-GeneChips. We  will add new chips covering microarray data for cereal species, such as wheat or rice,  when they become available. We regard this diversity in microarray data as essential ingredient for comparative genomics analysis at expression level. We have no plans to store spotted microarray data at this point.

1.4 How is BarleyBase integrating expression data with other genomic data?

BarleyBase is collaborating with PlantGDB and Gramene. This allows BarleyBase users to perform EST alignments and gene prediction using the Barley1 GeneChip exemplar sequences or cross-species comparison at the genome level. We plan to add links to dictionaries of cereal terminology and other databases in the futures. If you have suggestions, please contact the BarleyBase development team.

1.5 I find some vocabulary not familiar, where can I find their exact meanings?

For the experiment submission process please see the help fields for each field, as well as the  help at BarleyExpress pages. Information on the different types of GeneChip normalization and data processing is available on the data set query page. For some Affymetrix GeneChip-related terms, you may find it at glossary.php.  We will add a terminology page in the near future.

2.1 What's the submission interface for BarleyBase?

BarleyBase features a web-based experiment submission tool, BarleyExpress, which allows users to manage their experiment descriptions, array design, and expression analysis information, either individually or as a group. This interface organizes the experiment by experimental factors and defines a organized submission table.

2.2 Is the  submission tool, BarleyExpress, consistent with the  MIAME standards?

BarleyExpress is our MIAME-compliant online submission tool. It is based on ArrayExpress, the MIAME-compliant submission tool developed at EBI. (MIAME-Minimum Information about Microarray Experiments).

2.3 What is needed before I start my submission?

The basic requirements for making a submission are:

(1). Registration with BarleyBase as an official user. The process is relatively painless.
(2). You experimental design.
(3). Your protocols.
(4). The raw data from Affymetrix MAS 5.0 Suite. *.DAT and *.CEL files are required for each hybridization. *.EXP and *.CHP files are recommended.
(5). A browser with Sun java virtual machine (JVM) version 1.4.1 or above enabled.

2.4 Which operating systems can be used for making submissions?

Windows, MAC and Linux all work, provided you have a compatible browser installed.

2.5 Which browser works best for the submission process, how do I set my browser to make a submission?

BarleyExpress works best under Internet Explorer and Mozilla. For MAC, Safari is recommended. All browsers need the updated version of java virtual machine (JVM) for batch uploading of large amounts of data.

2.6 Why does BarleyExpress require experiment factor and factor levels?

BarleyExpress needs the experiment factors and their levels explicitly specified. We think this is essential for other users to understand the experiment, as well as helping  BarleyBase to analyze and present the data in a more meaningful and easy-to-understand way. As more experiments are added into BarleyBase, this will also help users to compare results across experiments.

2.7 Must I finish  a submission at a single time? Can I stop at some point and resume submission later?
Yes, you may stop at any time point and resume submission later. However, we delete all unlinked data files from our FTP respository after they have been stored for a week.

2.8 Can I add more experiment factors or factor levels to my finished submission?

No, this is one of the few fields that cannot be changed without resubmission. You can rename fields and values however (see section 2.9)

2.9 What I  change the names of the experiment factors/level?
Changing the names of factors or factor levels is allowed before finalizing your experiment submission, provided that this change will not change the numbers of factors and levels.
 

2.10 Any recommendations for making a smooth submission?

 Submission requires careful checking before presenting your results to other researchers in BarleyBase. Based on experience, some things to keep in mind for making a successful submission are:

1. Specify the number of factors and the levels of each factor carefully. Once submitted, this information can't be changed without re-doing the entire submission. The only allowable modifications are typo-corrections or renaming the factors and levels.
2. For the number of replicate per treatment, specify the maximum number per treatment you may need. This number can't be increased once submitted.
3. CAREFULLY check your file name association and sample association with the treatment before finalizing your submission. Complete any fields in the "Sample Preparation" page that you have information for.
4. Any revisions, except for text changes, after "Finalize" will incur much effort at both sides from submitter and administrator, and greatly delay the processing of your data in the best situation. Significant revisions may require re-submission.

2.11  I have many samples, and they are very similar, how to submit them quickly?

 Submission sample preparation information is the most time consuming part for experiment submission. Our solution is to let user copy existing sample preparation for editing. This means user will need to just fill in one sample preparation completely, and for all subsequent samples, user can simply copy this "template" sample, changing name and several other field that differ from your sample "template" . Using this strategy, sample preparation submission can be accomplished in hours, even for up to 100 samples.

3.1 Why register with BarleyBase? How to register with BarleyBase?

Only registered users can submit their experimental data, access non-public data, and perform batch downloads of data.. Registration can help us keep you informed of important improvements and new functionality in BarleyBase.  Registration is the prerequisite for data owner to grant privileged access to their data before the data is made public to all BarleyBase users.

3.2 What's the purpose of group management?

Group management lets submitters grant data access to their collaborators before the data is made public. For example, they may let manuscript reviewers access the data. They may also share data with consortium members worldwide. Group management  offers users the tool for achieving this. 

3.3 Will my submission be public immediately? Can I control who will be able to access my submission before my results are published?

BarleyBase encourages data sharing with the whole community. The default setting on all submissions is public. But we understand your need for keeping data private until your results are published. To make your data visible ONLY to users that you select, please designate your submission as "Private" or "Group Consortium" at when you submit your experiment design in BarleyExpress. Then you can use group management to grant access to people you select.

3.4 How can I grant access to my private data to other users? Can I  delete users from the access list?

Under the "My BarleyBase" menu, select "Group Management", then you can create, edit, or remove groups. You can also add or delete users from your own groups.

To grant access to an experiment to a given group, under the "My BarleyBase" menu, select "Experiment Management", follow the link Associate groups with the experiments.

3.5 What if I designated my data as private when making my submission, but later on I need to let selected people to access it?

You can create groups, and associate experiments with the groups as explained in 3.4.  

3.6 Can I get all data for an experiment with batch download? What types of data will be downloadable?

Yes, we make all data in compressed format for convenient batch downloading. There are four types of batch download:  raw *.DAT files (images read by Affymetrix scanner), the raw *.CEL files (Probe-pair level information), the normalized expression value files (using MAS 5.0 and RMA methods), and *.CSV filed generated by combining a CEL file with the corresponding CDF file.

[Back to Top]


4. MICROARRAY EXPERIMENT AND PROBE SET QUERY

4.1 How can I find a particular experiment?  Can I view all experiments stored in BarleyBase?

You can query for experiments on the "Experiment Query" page. Only public experiments, and the private experiments whose owners grant you access to will be available for you to query. The same set of experiments are also open to you for probe set query.

4.2 How can I retrieve data for my own list of probe sets?

You may do it at query_probeset.php. This is useful for users obtaining their interested set of probe sets through other strategy.

4.3 Can I locate probe sets that may be related to a given gene name?
Most exemplar sequences have some annotation. This is compiled from different sources by BarleyBase as part of our comittment. You may search for keywords within the description. Please follow instructions on the page.

4.4 How does a pathway or gene family behave in response to treatments from a specific experiment?
For Arabidopsis, you may browse the list of pathways or gene families, then select an experiment, and view the expression and annotation for all probe sets related to your selected pathway or gene family. The gene family information is compiled and made public by TAIR. The pathway information is provided by KEGG.

For Barley, such information is not available yet. You may try get more annotation detail for the barley sequences from Gramene, or PlantGDB.

4.5 I have sequence from other species, how to find its homologous exemplar sequences and their expression in the GeneChips?

The best place to go is the "BarleyBase Blast" page. Simply input your sequence and BLAST. From the results page, pick your  exemplar sequence-of-interest for a detailed view.

4.6  How to find probe sets meeting my criteria on expression profile?
The "Microarray Probe Set Search by Expression Profile" implements nine different data filters, which can be use alone, or in combination. They include common filters such as fold-change, variance and p-values. Basic understanding of single-channel microarray data from Affymetrix GeneChips will be helpful in choosing useful filters and options according to your needs.

4.7  My search returned a list of probe sets, can I get the probe-level data for them?

After probe set query, click on the links with the probe set name, and your will be at the detail page for the probe sets. At this detail page, click "Probe-Level Intensity" link you will be presented with the PM and MM intensities for probe pairs from the probe set. Try out following the Contig3_at or  Barley1_00003  links in 4.9  to get to probe-level data.

4.8  I would like to analyze a data set obtained from my queries locally, what format is the data in?

You may now download them as tab-delimited text files, either as expression, or as annotation. This plain-text format can be easily imported to most microarray tools. We will add more download formats such as MAGE/ML in the future for offline data analysis.

4.9   Where to get all information for a given exemplar sequence or probe set ?

Detailed information for a Barley1 exemplar can be retrieved with its ID, such as Barley1_00003. This page listed everything available within BarleyBase about the exemplar.  Users usually access such pages through link in probe set query pages, or through "Probe Alignment" pages.

For details about a given Barley1 probe set, you may reach it by following links at the probe set query result pages. In addition, each exemplar detail pages has links for corresponding probe sets too. Try  getting from Barley1_00003 exemplar to Contig3_at  probe set.

For ATH1-121501 GeneChips, you may view detailed exemplar annotation at TAIR  for the genomic locus following links at the probe set query result pages.

[Back to Top]


5. DATA ANALYSIS AND VISUALIZATION WITHIN BARLEYBASE

5.1  Can I analyze the data sets from a query using BarleyBase? If so, what analysis is available?

BarleyBase provides most types of microarray data analysis methods,  including hierarchical clustering, K-Means and k-medoids partitioning,  self-organizing maps (SOM),  principal component analysis (PCA) and Sammon's Multidimensional Scaling (MDS), differential gene identification.

Data sets obtained from probe set filtering or queries are integrated with microarray data analysis.  Follow the "Save/Download/Analysis" button on filtering result page to perform customized analysis.

Or you may analyze saved data set directly at analysis page.

5.2  Can I select a sub-cluster from the dendrogram to do more detailed analysis?
You can specify the number of desired sub-clusters at the analysis set up page, and each sub-clusters will be available for refined analysis by saving it to a new data set.

We will add functions to let user select probe sets directly from the dendrogram with mouse when time permits.

5.3  How is the online data analysis carried out at BarleyBase?
We are using R and Bioconductor as the backend for our data analysis. Bioconductor is a open source effort for building microarray data analysis tools in R, an open-source statistical software package.

5.4  Is data transformation and normalization important? How to decide when this is essential?
Most of the analysis use the dissimilarity matrix in some way. The dissimilarity matrices usually highly depends on the scale of input data. So you can expect the results may vary if the scale changed.  Thus log2 transformation will always have drastic impact on results.

Data centering will affect most analysis too. One exception maybe the correlation-related distances, where only the relative change matters.

There is no gold-rule that can be applied under all situations.  You need to consider the analysis you are to conduct, the data structure and the distance measures you need.  An in-depth guideline on this issue is out of scope of this page. At least, you may want to run analysis under different settings and compare the consistence across runs.

5.5  What types of microarray data visualization is offered in BarleyBase?

Visualization is available at all levels of data, from experiment to probes. The full detail is at Visualization page.

1. At the hybridization level, or at treatment level, scatter plots and M v A plots are provided to aid user in finding the expression changes across two hybridizations or treatments. This is accessed from most probe set query related pages under "Visualize" link.

For each hybridization (chip), histograms for the perfect match (PM) and mismatch (MM) probe intensities and the image for PM intensities are available though the "Hybridizations Details" pages. Just click on the hybridizations to view them.

2. At probe set level, multiple visualization methods are provided.

(1). After prove set queries, each probe set is associated with the line-plot through  "View" links under "Expression Plot" column and the heatmap-style expression graphs with all profile neighbors through  "View" links under "Profile Neighbor" column. Both types of graphics are iterative.

 (2). After query and analysis, all sub-clusters or sub-partitions of probe sets are depicted by expression graph and heatmap graph. These graphics are static.

3. At probe level, bar plots for PM and MM signals for probes from the probe set are drawn by hybridization and by probe pair numbers. At the detail pages for exemplar or probe sets, follow the "Probe-Level Intensity" link, select an experiment, click the "Get barplot for xxxxx." link to draw bar plot for the selected experiment. One example is here: Contig12438_at.

Last Modified: 10/02/2003