<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 3.2//EN">
<HTML>
<HEAD>
<META HTTP-EQUIV="Content-Type" CONTENT="text/html; charset=iso-8859-1">
<META NAME="Generator" CONTENT="MS Exchange Server version 6.5.7654.12">
<TITLE>RE: [ome-users] .mrc file import into omero</TITLE>
</HEAD>
<BODY>
<!-- Converted from text/plain format -->
<P><FONT SIZE=2>Some background and few comments about emdatabank.org and the mrc format.<BR>
<BR>
EMdatabank is collaboration between the cryo-EM community, the PDB and EBI, going back to 2002 when a declaration was made to encourage publishers to require open access to datasets that authors reference in their submissions.<BR>
<BR>
The EM Community Outreach History is here:<BR>
<A HREF="http://emdatabank.org/?q=node/68">http://emdatabank.org/?q=node/68</A><BR>
<BR>
The baseline metadata specification, comparable to the OME specification, can be found here:<BR>
<A HREF="http://emdatabank.org/?q=dictionary">http://emdatabank.org/?q=dictionary</A><BR>
<BR>
EBI hosts a web page with ccp4 and mrc map format definitions:<BR>
<A HREF="http://www.ebi.ac.uk/pdbe/docs/iims/3D_EM.html">http://www.ebi.ac.uk/pdbe/docs/iims/3D_EM.html</A><BR>
<BR>
<BR>
<BR>
The MRC format is arguably the oldest biological digital image format, and predates TIFF by several years.<BR>
<BR>
The MRC format, a/k/a map format, was derived from the MRC software suite used in high resolution EM image processing, dating to ~1969.<BR>
<BR>
The file format was added to the suite around 1982 as an ad hoc format similar to the philosophy of the MRC suite. <BR>
<BR>
The xray community picked it up a few years later in their formal standards activity, adding their metadata and calling their format the CCP4 map format.<BR>
<BR>
At this point there was a loose relationship between the EM community and the X-ray, a common intersection being electron crystallography; but generally the formats drifted apart.<BR>
<BR>
Around 2000 there was an effort to harmonize the formats, which has generally held.<BR>
<BR>
Each header is demarcated as 256 32-bit words (1024 bytes).<BR>
Within the header there are a number of undefined words, 38-52 are defined as "future use". <BR>
<BR>
Within the ad hoc EM community the philosophy has been from the beginning these are not reserved and may be used for any purpose the user or software developer chooses.<BR>
<BR>
A significant complication is that neither format establishes versioning information, so sorting things out can be a problematic and mysterious.<BR>
<BR>
Another major difficulty is the EM ad hoc metadata addition of "origin" (presumably the lowest corner of the bounding box) at words 50-52, which differs from the CCP4 ad hoc definition for another variable. This is a critical problem, because a number of visualization programs use this for graphics in the EM community.<BR>
<BR>
Within the IMOD program, it uses a number of "future use" words for various purposes.<BR>
<BR>
<BR>
We are currently in discussion with OME/LOCI to resolve this problem, as well as a number of other image format problems, systemically.<BR>
<BR>
Besides the metadata issues, there is the looming problem regarding the quantity of pixels, which can make the image intractable. <BR>
<BR>
For example in my lab we are planning for a 10k^2 CCD, which will result potentially in 512GB 3D images.<BR>
<BR>
If any one has suggestions, comments or wants to know more, please contact me directly.<BR>
<BR>
<BR>
Matthew Dougherty<BR>
713-433-3849<BR>
National Center for Macromolecular Imaging<BR>
Baylor College of Medicine/Houston Texas USA<BR>
=========================================================================<BR>
=========================================================================<BR>
<BR>
><BR>
> As has been mentioned, we are planning to support EM images in OMERO <BR>
> in a project spanning the next 3 years, in a collaboration with the <BR>
> EM DB at EBI.<BR>
> This includes building a complete data model for EM data (as we have <BR>
> done for light microscopy) and implementing this in BioFormats, the <BR>
> OMERO server and clients.<BR>
><BR>
> As the OMERO team has little experience of EM to date, any help that <BR>
> you can provide in this area would be most welcome!<BR>
<BR>
Quentin is the guy here wgho is involved and interested in image <BR>
database solutions for electron microscopy,<BR>
and i wonder if he might like to help you define the EM meta data model.<BR>
<BR>
He has a much better i dea than i do about that meta data is really <BR>
important for EM.<BR>
<BR>
cheers<BR>
<BR>
Dan<BR>
</FONT>
</P>
</BODY>
</HTML>