The Hard Work
(and how it was done)

History of the Package

The MDPP was written by P.R. Smith on the F. Hoffmann-LaRoche IBM 360 & 370 computers in Basel, Switzerland, beginning in 1972. Additional program was written by Drs. U. Aebi and A.C. Steven. MDPP was ported in 1978 onto the PDP 11/70 at NIH by Drs. Benes Trus and A.C. Steven. The NIH system (re-named PIC) was given to U. Aebi at Johns Hopkins in 1980 who passed it back to P.R. Smith at NYU at the end of 1981, with improvements incorporated by Dr. Loren Buhle in the Aebi lab, where it was merged with the then current IBM version of the MDPP to create a PDP11/RSX11M version of the MDPP. Since the transport of PIC to JHU, the MDPP and PIC systems have diverged, and are now distinct packages with a common root in the old MDPP/IBM from Basel. The first version of MDPP/VAX was generated by Loren Buhle at JHU based on MDPP/RSX, and was brought to NYU in 1985 for debugging and reorganization. The port to Ultrix was also started by Loren Buhle, and was finished by Ross Smith (in 1991). Rachel Oberai at DEC supplied a DECstation 3100 to use as a conversion platform. The OpenVMS/AXP port was performed in NY in March, 1993, also after the loan of a computer from DEC. The NYU Medical Center GCRC provided the use of the Alpha-DEC UNIX platform for the port to that operating system. The Kaplan Comprehensive Cancer Center provided support which enabled us to port the software to the SUN and SGI platforms.

Many other individuals have contributed code incorporated into the MDPP, including (in no particular order) David Agard, J-P Bretaudier, Joachim Frank, Terry Frey, Michael Radermacher, Owen Saxton, Michael Unser, Lyn ten Eyke, Tony Crowther, Linda Amos, Robert Solomon, Billy DeLeon, Terry Peters, Peter Maas, Suzy Gottesman, Roland Buerki, Tim Baker, Richard Henderson, Mike Schmid and Andreas Engel. Marin van Heel and Michael Schatz helped with the movie program and provided hints to aid our X-programming. Tom Clarke did most of the work getting the on-line BOOKREADER documentation started. A stripped-down Cray version of the system was developed at PSC in 1990 with the hard work and skill of Doug Balog and Jim Joyce, student programmers, and Alex Ropelewski and David Deerfield on the PSC staff. The HTML version of the documentation was built by Ti Tang.

This version of the MDPP has been extensively rewritten since publication of a short report on the system in 1978 (P.R. Smith, Ultramicroscopy, 3:153), including a completely new user interface. One major strength of MDPP is that it has been constructed to make use of compute-servers for the analysis of large computational tasks. The MDPP running on a small workstation (under VMS or Ultrix) will decide whether the computational task assigned by the operator can be done locally, or whether a remote compute-server could provide better service. MDPP then selects the appropriate compute-server, either the VAX 6000 front end or the PSC Cray Y/MP, and automatically sends the job to the server. The operator either remains on-line to watch the progress of the job (in the case of the VAX 6000 as server which employs an active DECnet link), or is free to do other tasks until notified of the completion of the remote job and the availability of the computed data (as with the Cray). The description of this is published in CABIOS, 7:501-507, 1991.

The most recent description of the MDPP is to be found in the newly-published special issue of the Journal of Structural Biology (P.R. Smith and S.M. Gottesman, JSB 116 Number 1).

Availablility, Copyright & Support

Availability

MDPP is distributed electronically at the cost of tapes and labor (if any). People who have MDPP are free to re-distribute it to other research scientists, provided they do so at no cost other than that of the media and postage, and fully inform the recipients of the conditions of the Copyrights on the software and the limitations that have been placed upon its use.

Copyright

Please note that the MDPP package is covered by US Copyright. It has been provided for scientific research only and is not to be used for commercial purposes. Please note that the removal of the Copyright notice in the source code is illegal, as is the distribution of the package, or any part of it, without the Copyright notices. Additionally, the MDP program may not be renamed; the announcement which appears when the program starts MUST contain the words `BIOZENTRUM MICROGRAPH DATA PROCESSING PROGRAM' and the Copyright notice. The Correspondence Analysis package has been provided by Dr. J-P Bretaudier. It is also covered by US Copyright and has been provided for scientific research only. It is not to be used for commercial purposes. Please note that the removal of the Copyright notice in the source code is illegal, as is the distribution of the CORAN package, or any part of it, without the Copyright notices.

Acknowledgment of Support

The development of the MDPP represents a huge investment by the authors, and also of the funding agencies that have helped pay the development costs. The major contributor recently has been the National Science Foundation (NSF) which has contributed indirectly and directly through grant BIR-9410750.

It is, therefore, a condition of use of the MDPP that the support for its development as scientific research tool be acknowledged in all publications describing the work. An example of a suitable phrase is: "The development of the MDPP package was supported in part by the NSF through grant BIR-9410750" but modifications of this may fit better in some cases. For the purposes of defending my grant I would be very grateful if people were to E-mail me with the citations of papers that describe the science done with the help of the MDPP.

I understand, of course, that everyone tries to minimize the clutter in the `Acknowledgments' bit in the paper. It has been made quite clear to me by the NSF, however, that in these tough times, NSF, NIH and indeed all US Federal agencies supporting science need to justify their activities by showing a product. The best product is translational science, that is, science that starts in a lab, and turns into a deliverable like a cure for the common cold, a new software package to replace Lotus 1-2-3R, or whatever. Failing this, lists of papers documenting that the Agency investment produced something that others cared about, is an acceptable alternative. Showing use of tools supported by development funds is therefore essential to preserve these programs. Please help!