Symbols and icons
| Symbol | User action | Description |
 | click | More information.
Click to read the experiment protocol, and to see info on publications/investigators, as well as
a units/abbreviations glossary. |
or  |
click | Predefined comparison view involving two or more measurements |
| ▲ or ▼ | see | Significantly higher (lower) than overall norm
in measurement lists with strain data displayed. Red=female, blue=male. |
| ◀ or ▶ | see | Left (right) side significantly greater
in measurement lists with female vs. male or two strains being compared. Red=female, blue=male. |
 | click | Plots & tools wizard. Begins the
wizard using measurements currently listed, plus any flagged measurements (see below). |
 | click | Data downloads |
 | click | Flag (add) a measurement to your collection (described below) |
 | see | Already flagged indicator |
 | click | List correlations.
Scan all MPD for measurements that are correlated based on strain mean data points |
 | click | SNPs or to deploy the "find genomic regions" SNPs tool |
Keyboard, mouse, and browser hints
- The
symbol appears in many places... click on it
to consult the experiment protocol, to see participants or publications, or to get further info.
- Our web site uses "tooltip" text bubbles in many places.
Try pointing at areas on the screen (without clicking) to see additional information or hints.
- Our web site frequently deploys popup windows (always from phenome.jax.org)...
for best results turn off or adjust browser pop-up blocking.
We utilize browser cookies for user preferences, flagged measurements, and in a few other places,
so cookies should be allowed from phenome.jax.org.
- To select more than one item from a multi-choice selection list, hold down the Control (or Apple) key while clicking.
- To control the deployment of measurement graph windows see Preferences.
Otherwise, you can usually hold down control (or Apple) key while clicking on a hyperlink in order to deploy the new content into
a separate window or tab, useful for on-screen comparison.
- MPD is developed (and displays optimally) on Firefox.
We recommend that MS Internet Explorer users use MSIE's "smaller" text size
(use menu selections: View / Text size / Smaller)
- Browser bug (Safari, Chrome) ... clicking/pointing to plots and images when in zoom page mode --
interpreted location is incorrect. Workaround is to be sure to not use zoom page mode when working
with plots that have point/click functionality.
The bug is discussed here.
Posted 29 Oct '09
Preferences
A number of useful settings can be made by clicking on the Preferences link at the top of any MPD page.
You can control the format of measurement listings (compact vs. spacious), the home page subject area
list (succinct vs. expanded), whether to deploy measurement plots into one or many new windows, and
a variety of appearance-related preferences.
What is an MPD "project"?
We use the term "project" to refer to a data set
contributed by an investigator (or sometimes downloaded from a public resource).
Most MPD projects contain phenotype data but there are some
genotype projects (primarily SNPs) as well.
MPD is an open community project and we are always looking for good data sets. If interested
in contributing data see further information under "About MPD".
Phenotype projects contain data captured under controlled conditions
using well-defined protocols. The results are organized into groups of "measurements".
The measurement data are available for online analysis or download, as are the
protocol documents. Projects are usually "authored" by one or more investigators,
and are conducted within a months/years timeframe (MPD displays the year that labwork was conducted).
Projects usually involve a common
cohort of animals, and are often connected with peer-reviewed journal article(s).
Sometimes very large phenotyping efforts end up as several MPD projects
for ease of handling and navigation.
Projects are identified using a "project symbol" which is the principal investigator's name
followed by an integer, such as "Smith1". Projects may also be identified by accession number
such as MPD:29.
What is an MPD "measurement"?
There are hundreds of phenotype "measurements" or "traits" stored in MPD.
A project (see above) can include anywhere from 1 to 80+ measurements.
Each measurement is a test that has been performed on mice of various strains.
Here's an example of a measurement:
The top line (yellow) shows the primary subject area.
The second line includes the project symbol (Deschepper1; click on it to see more about that project).
Next is the measurement description (click to open a plot and summary of the measurement),
followed by number of strains tested and animal age at test time.
Data for females and males is always segregated in MPD.
Most measurements are done on both sexes, but for those that involve just one
or will be shown at the far right.
You can always click on to get more information and jump to the experiment protocol text.
Clicking on will flag the measurement (discussed below).
Here's another example:
If the measurement involves an intervention, such as special diet or drug dose, it is indicated (often in an orange box
as shown here).
Where appropriate, related measurements are displayed together as a "block", as in the above example.
This block includes 3 measurements of non-HDL (baseline, after highfat diet, and amount of change after diet)
from the same project (Paigen2).
Again, you can click on to get more information, and
clicking on will flag the whole block as one item (discussed below).
Click on the item to see a graphical comparison
(in this case comparing before/after diet). Predefined comparison views like this are provided wherever it
makes sense to do so.
In this example, note that a special strain panel has been tested, denoted by the green-shaded box.
If no special strain panel is mentioned then inbred/general can be assumed.
Useful comparison views that involve measurements from different blocks are made available as individual entries and
denoted using the symbol. These are not "measurements" per se.
For selected traits where inbred strain data has been contributed from multiple sources, MPD offers
predefined "Composite/Consensus" views which allow easy viewing and comparison of data from multiple sources (added Sep '07).
A list of all available Composite/Consensus views is available via the home page.
How do I find particular phenotype measurements?
There are several approaches:
- Start at the MPD home page and browse the categories listed there (try "Expand list" which allows you to work with
multiple categories together if desired).
- Do an MPD search from the home page.
- Do a Google MPD search (see link on home page).
- If interested in the work of a particular investigator, browse by project/investigator from the home page.
You can view one project or several together.
- If interested in particular strain(s), browse by mouse strain from the home page.
I've retrieved a list of measurements for "blood lipids". How can I see values for DBA/2J, or
see comparisons of DBA/2J vs. C57BL/6J?
Scroll down to the end of the list, where there are some interactive controls for
displaying strain mean values. Female values will be shown in red, male in blue.
The little triangles are explained in a section below that listing.
How can I see overall averages (all strains combined) in the "blood lipids" category?
With the "blood lipids" category visible, scroll to the end of the list, where there is
a control for doing this. You'll see overall averages, along with
± SD which gives an indication of the amount of variability across strains.
Suppose I have found some data of interest but
I don't see a clear, direct answer to what I am seeking. What do you recommend?
If it's a problem with presentation format or notation, be sure to look over the entire
web page... often additional explanation is provided at the bottom of the page. Questions
and suggestions can be sent to us at phenome@jax.org
Interpreting the data:
MPD curates and posts data as contributed by investigators from their experiments.
In some cases results are scalar and clearcut; in other cases there are
multiple variables and other complexities that require
significant delving and analysis on the user's part.
MPD currently does not provide high-level interpretations of data, but
if a publication is associated with the data this is often a
good place to find relevant conclusions and discussion. Another option is to
open a dialog with the contributing investigator.
Find protocol, publication and investigator info by clicking on wherever it appears.
Suppose I have found some data that seem to answer my question.
What do you recommend with regard to "due diligence"?
We recommend consulting the experiment protocol and publications
to be aware of nuances and details that might have an impact on the data.
Find protocol, publication and investigator info by clicking on .
If more than one investigator/project has provided relevant data in a certain subject area,
all available data should be taken into account.
Of course the usual statistical power considerations also apply.
What if I encounter data that seem contradictory or not in agreement?
This can occur. MPD does not act as referee on contributed data.
We encourage different labs/investigators to contribute data in all subject areas for validation,
redundancy, and consensus-building purposes. However, this as well as MPD production of
consensus/composite views is a long-term goal.
Are values rounded? How does MPD decide what level of numeric precision to present?
When MPD accessions numeric data, each value is automatically rounded to a certain precision based on
the magnitude of the individual number.
In order to avoid loss of information the rounding algorithm often provides a bit more precision than is necessary.
This method sometimes results in different precisions for two similar numbers when one is over and the other
under a certain magnitude cutoff.
MPD does not attempt to control the degree of numeric precision on a measurement-by-measurement basis, therefore
the MPD-displayed precision may not be the same as the precision produced by a given laboratory apparatus, or
as supplied by an investigator.
What statistical processing does MPD perform on phenotype data?
When MPD accessions numeric data,
summary statistics for each strain / sex / measurement are computed and stored, as well as summary statistics
on the overall measurement (such as overall male mean and overall female mean).
Female vs. male comparisons (with t test p values) are computed and stored for each strain / measurement.
Pearson correlation coefficients (with p values) for each incoming measurement vs. all other MPD measurements are also
computed and the stronger correlations stored, for purposes of mining for correlated measurements.
Other statistics are computed on the fly when web pages are produced, including
comparisons of two strains (with t test p values) and Pearson correlation coefficient matrices.
What data viewing and analysis tools are available? How do I use them?
As mentioned above, if you're looking at a category list (or other measurement list) you can
easily list strain data, sex differences, or compare two strains using the controls at the end
of the list. The predefined comparison views have also been discussed above.
Additional display and analysis tools are available via the
Plots & tools wizard. This can be invoked
from most pages where measurements are listed. Here's the procedure:
1. Start at any desired phenotype subject area list, or other list of measurements.
2. Click on Plots & tools wizard.
3. Select measurements by clicking on checkboxes. If you've flagged any measurements
these are also available for selection. Then click on Next.
4. Select the plot or tool you want by clicking on the appropriate graphic icon.
5. See your result.
6. Click on "Start over wizard" to repeat the process.
If you just want to see the array of plots and tools that are available,
go to the home page and click on "Plots and tools preview".
What are "flagged measurements"?
These are MPD measurements that you gather into a personal collection.
This is a good way to "bookmark" measurements of interest for future reference. It also allows you to
get measurements from different categories together in one place for use with plots and tools.
To flag some measurements, click on the symbol for every measurement/block you wish to collect.
Your collection is stored using a browser cookie (called MPD_FAVMEAS) that will be available to future MPD sessions.
You can collect up to 40 items at a time. Click on the "Your flagged measurements" link at the
top of any MPD page to see your collection, or to use, edit or completely clear it.
How do I use the Plots & tools wizard with measurements from different categories?
From the home page, click on "Expand this list", select the desired categories using the checkboxes,
then click on "Plots and Tools wizard" (as of Sep '07).
You can also do it the old way —
gather the measurements of interest into a collection (see above)
then invoke the Plots & tools wizard.
How do I see everything MPD has on a specific strain?
From the MPD home page, under "by mouse strain", click on "select a strain".
This will take you to an MPD page for the strain of interest, which contains subject area links
to all data in MPD for the strain, as well as links for viewing SNPs.
How do I compare two specific strains (phenotype or genotype)?
From the MPD home page, under "by mouse strain", click on "compare 2 strains".
You will reach the MPD two-strain comparison center page for the two strains of interest,
which contains subject area links
to all data in MPD involving both strains, as well as links for viewing SNP comparisons.
Another way to compare two strains for a specific measurement list is to use the controls
that are located below the list (mentioned above).
How can I find mouse strains that best meet a certain phenotype criteria?
Invoke the Plots & tools wizard, and select the "Find strains that meet your criteria" tool.
How can I acquire or download MPD data to manipulate locally or load into a local database?
- For capturing a table you currently have on your screen, interactive copy/paste is convenient (see below).
- To build your own data set of selected phenotype measurements from various sources, use
the custom data set tool via the Plots & tools wizard... you can get
strain means/summary stats or raw animal values, and select strains if desired.
- For SNPs go to the SNPs page.
- To download one project (one contributed data set) go to the appropriate project page (Home / By project).
- Otherwise, for bulk downloads (entire MPD database tables) and all other available downloads,
click on the Downloads link at the top of any MPD page, which takes you to the MPD download center.
Can I copy/paste a table from an MPD web page into Excel or MS Word?
YES. Most MPD data tables can be copy/pasted directly into Microsoft Excel for local manipulation and analysis.
This works with MSIE and Firefox browsers, and maybe others:
1. For best results set black line table borders preference first. Click on Preferences at the top of any MPD page.
After you've set the black lines preference, refresh the page that has the table on it.
2. With the desired MPD table showing in your browser, select the table by dragging downward over it with the mouse.
Be sure to begin dragging the mouse above the top/beginning of the table in your selection.
Then on the top menu bar choose Edit/Copy (or hold right mouse button down and select Copy).
3. In Excel (or MS Word) select Edit/Paste. The table data should appear. If it doesn't come out right the first time, try once or
twice more using different starting points for the mouse drag operation.. there sometimes seems to be a bit of voodoo involved!
Can I open flat files in Excel?
Sometimes MPD offers a tab delimited flat text file but no Excel option.
YES. Save the flat file onto your machine.
If the flat file is zipped, use Expander (or similar) to unpack it.
Then start Excel and open the file... Excel will take you through a brief wizard process,
simply choose all the defaults and it should be fine. Keep in mind that Excel has some capacity
limitations beyond which Excel may freeze up or crash.
How can I capture MPD content for slides / PowerPoint?
One way is to do a screen capture. First maximize your browser window (or center it).
Windows users can press the [Print Screen] key, then in a PowerPoint blank slide do Edit / Paste.
Mac users can hold down [Ctrl] [Apple] [Shift] [4] (all four keys), rubberband the desired region,
and then drag it into PowerPoint.
You can also capture individual graphics/plots. Windows users can point to the graphic,
click the right mouse button, choose "Save picture" or "Save image as", and save the image onto desktop;
then in PowerPoint, do an Insert / Picture.
Mac users can simply click on the graphic and drag it into PowerPoint.
How do I get publication-quality graphics? How can I tweak the graphics?
Many of our graphical tools allow you to select between PNG, EPS, and SVG file types.
PNG is the default but isn't highest publication quality.
EPS is higher quality and can be handled by Freehand and Adobe tools (graphical elements
can be manipulated using Adobe Illustrator).
SVG is another vector format with the potential for excellent publication quality.
Is it ok to spider the MPD web site?
How do I report a problem? A certain MPD web site tool is giving strange results.
If you notice something that seems wrong or odd, we appreciate hearing about it at phenome@jax.org so that we can check into it.
Can existing MPD phenotype data change over time? How can users find out about changes? What about tracking software bugs?
Data sets presented on the MPD web site are not necessarily static. Corrections or additions
may be made to them over time.
It is also possible for invalid or incomplete results to be presented on the MPD web site due to software bugs, data problems,
or human error. Data corrections and bugfixes are generally recorded in the
MPD change log, and we recommend that this log be reviewed periodically.
Other questions or suggestions?
Contact us at phenome@jax.org or use the "Suggestion box" link at the bottom of
any MPD web page.
|