Announcments and Bugs

OSX Support Cancellation

Apple's diminished interest/support for scientific and commercial computing is obvious with the cancellation of Apple Server and its new OS update policy.

We ceased releasing SPIDER/Web for OSX in 2014.

New operations

New SPIDER operations or changes in usage of old operations.

Note that these changes may NOT be in the release of SPIDER that you have at your site now.

---

Recent Bugs

CA S

??? -- > 25.00     Bug in Iterative PCA Multivariate statistics.

AP SCC

23.00 -- > 24.00     Sub-pixel accuracy lacking when using Intel compiled SPIDER.

FQ & FQ NP

23.00 -- > 24.00     Volume filtration failed.

PJ CYL

??? -- > 23.00     Usually crashed.

AP REF

??? -- > 22.10     Sometimes gave nonsense alignments.

General

??? -- > 22.08     SPIDER did not support images/volumes with any dimension > 10,000, thus large MRC stacks would not convert to SPIDER volumes.

IN S

??? -- > 20.15     Scaling sometimes failed.

CTF FIND

??? -- > 20.12     Reports incorrect SPIDER astigmatism angle.

CP TO OPEND

??? -- > 20.07     Fails to create proper new stack header.

CP FROM MRC

??? -- > 22.11     Does not 'FLIP TOP/BOTTEM' on 32 bit volumes (works on images).

Does not 'FLIP TOP/BOTTEM' on 16 or 8 bit images or volumes.

FQ and FQ NP

20.16 -- > 21.03 Boundary used for determining whether units were in in pixels or frequency units was 0.5 which gave incorrect values if user intended to use Nyquist frequency as a filter setting.

TF ED and CTF ED

-- > 21.02 Astigmatism calculation unreliable and sometimes crashed. use CTF FIND instead.

MA

20.00 -- > 20.11 Crashed sometimes when masking volumes.

ROT

20.00 Crashed sometimes if using quadratic interpolation.

AP SCC

20.00 Crashed if using selection file.

AP SCC

19.07 --> 19.11 The cross correlation values reported were not normalized in rare cases.

Executable for: spider_linux_mp_intel

? --> 19.09 This executable was for a 64 bit OS although our documentation states it was appropriate for a 32 bit OS.

AP REF

19.07 --> 19.08 The 'CHECK MIRRORED POSITIONS' prompt requires a 'Y' and does not accept legacy '1' response.

FS, CP

19.06 --> 19.08 Sometimes failed to copy last image(s) when used to access a whole stack file at once (e.g. stkfile@).

DOC RENUMBER

19.01? --> 19.06 Sometimes crashed if comment lines have a '/' inside.

OR SH and AP SHC

19.01 --> 19.06 For certain inputs the reported rotation angles were sub-optimal.

AP I

18.10? --> 19.06 Reference rings files created by this operation for larger images failed to work in 'AP REF'.

AP REF

2000? --> 19.04 The X shift value is off by one pixel for even dimensioned images if 'Check mirror is '1' (or 'Yes') and the optimally aligned image would require mirroring of the input reference image.

SPIDER for MPI

? --> 19.04 Document operations crashed with bad: ICOMM message.

CE ST

? --> 19.04 Asks for extra unused input when 'automatic' option is selected.

SPIDER for OS X

18.20 --> 19.01 Some users report that SPIDER for 'OS X' fails to start on their systems due to lack of 'stack' space. This appears to depend on 'OS X' release.

IQ FI

? --> 19.01 Failed to give correct response for existing inline images/volumes.

VM M

18.02 --> 18.20 The 'VM M' operation sometimes fails.

AP REF AP SH and most other 'AP **' operations

? --> 18.13 These operation have a 0.3-0.5 degree error in in rotational alignment assignment for angles very near to 0 and 180 degrees. This error is negligible for reconstructions but the my correction will make the reported cross correlation values of all angles slightly different from that reported in previous releases.

SY DOC

? --> 18.12 The 'SY DOC' operation sometimes gave 'NaN' for orthogonal/cubic symmetry values.

AP SCC

? --> 18.08 The 'AP SCC' operation sometimes failed to give sub-pixel alignment results.

CA SR

?--> 18.08 The 'CA SR' operation when used with output from Iterative PCA did not properly subtract averages when requested to do so.

AP SH

?--> 18.04 The 'BP 3F' and 'BP 32F' operations sometimes crashed when using stacked projections and 'CP' within a loop.

AP SH

17.04--> 18.03 The 'MO 3' operation with 'CYL' operation gave volume with bad background.

AP SH

? --> 18.02 The 'AP SH' operation sometimes gave nonsense when used with multiple processes and a limited angular search range.

AP SCC

? --> 18.01 The 'AP SCC' operation gave nonsense when used with simultaneous multiple images and multiple references.

AP SR, CC, ..

? --> 17.11 Certain operations involving correlation e.g. 'AP SR' sometimes crash when compiled for Linux using Intel compiler (e.g. for OS X).

AP REF

? --> 17.08 Register 10 ('Change in projection angle') of the alignment output file has incorrect values when used with multiple images.

Nextresults

? --> 17.08 Nextresults script missing from distribution. Gave error message on starting SPIDER.

TF ED

? --> 17.07 Often crashed when compiled to run on Intel processors (OK on AMD).

AP SH

? --> 17.07 A "TRANSLATION SEARCH RANGE" that goes out of image reports an error message but fails to completely stop a parallel operation, leading to incorrect alignments.

CL CLA

? --> Dendrograms do not display image class assignments. It starts at class level. A design decision not a bug.

AP SR

16.** --> 17.01 Sometimes gave all zero alignment angles.

BP CG

17.00 --> 17.01 Sometimes failed to find angles for projection. (Results were correct if it ran.)

CL CLA

? --> 15.12 Gave incorrect class assignments when certain numbers of factors requested. Also always used first factors even when factor list was specified.

? --> Mar. 2006, A BRIX volume converted from a SPIDER volume by using "CP TO BRIX" can be used to fit a PDB structure using "O" or Real Space Refinement (RSR). Usually, the coordinates of the PDB structure are changed to reflect its fitted position relative to the BRIX volume. The BRIX volume and the new PDB coordinates are in correct orientation and location with respect to each other. However, if the SPIDER volume used for fitting is now changed to a CCP4 volume by using SPIDER command "CP TO CCP4" then the resultant CCP4 volume will not be at correct location and orientation relative to the coordinates of the above fitted PDB structure. To put the CCP4 volume at the correct location and orientation with respect to the fitted PDB structure, one will have to use the SPIDER procedure "fit_ccp4_to_pdb.spi" (see CP TO CCP4). The inconsistency is due to different coordinate conventions used in SPIDER and "O".

Note :, Some of the earlier SPIDER volumes deposited at EMDB database by our group at Albany were converted to CCP format by directly using the SPIDER command 'CP TO CCP4' (SPIDER procedure "fit_ccp4_to_pdb.spi" was not used). These volumes are not at the correct location and orientation with respect to the coordinates of the deposited fitted PDB structures. We have resubmitted all these volumes with correct location and orientation which respect to the coordinates of the deposited fitted PDB structures. These volumes are now available under category "other" of the EMDB database download page.

03/9/06 BR

AP REF, CC & CC N

? --> 14.10 When running our precompiled executable: spider_linux_mpfftw_opt64 (only this executable), the output image (CC) or the translational shift (AP REF) may be WRONG. This was due to a complex arithmetic bug in the PGI compiler which we used. This problem has been fixed by rewriting the relevant code to avoid the use of COMPLEX variables.

02/14/06 BR and AL

RF & RF 3

--> Oct. 2005 (version 13.00) Gave low 3-sigma resolution

Number of voxels (fifth column in the output document file) too low by a factor of two and the 3-sigma criterion output (fourth column) was too HIGH by a factor of SQRT(2). This means that resolution is better than reported.

10/7/05 PP

Older bugs

Bugs in earlier releases of SPIDER.