Q: I have a heterogeneous data set and want to know where I can find information on the phase residuals of the data set in EMAN to see if it can reveal the extent of the heterogeneity of my data.

A: What do you mean by 'the phase residuals of the data set' ?

EMAN does not use actual phase residuals for anything, though, if you use the 'phasecls' option, it uses the related 'mean phase error' for some aspects of classification. I think what you are looking for is probably the similarity values between the particles and the reference images, which in EMAN is not a phase residual, but serves a similar purpose. However, I fear it will not be very useful to you in this context, unless you are doing a single defocus experiment without CTF correction, and have very uniform specimens. That is, normally, the variation in the similarity values due to changing defocus and image contrast will far outweigh differences due to specimen heterogeneity. If you were collecting all of your images under basically identical conditions, then you might be able to do something along these lines, however, I'm still a bit dubious that you would, for example, see a bimodal distribution in the image similarity values if you had a 2-state heterogeneous system. The EMAN1/FAQ/Heterogeneous page has much better ways to look at this issue.

If you do want to play with the similarity values produced during classification, they are stored in the 'particle.log' file. This is a 'tagged' text file, with the first character on each line being the type of data it contains. You want to look for 'C' lines:

C <ptcl #> <usu 0> <orientation info> <class assigned> <usu 1>

the <orientation info> is further divided into:

<similarity value>,<da>,<dx>,<dy>,<flip>

Where 'similarity value' is the number you want. It's meaning changes depending on the options you give the refine command. Hope that helps.

EMAN1/FAQ/HeterogeneousPhase (last edited 2008-11-26 04:42:29 by localhost)