Models used as input to modTwist2EM must satisfy several requirements.
When modeling microtubules, for example, individual protofilaments can often be traced for some distance in a tomogram, until they become indistinct due to missing tomographic information. In this case, separate contours would be created in the input model for each protofilament traced. Of course, it is also permissible to break contours even when not changing protofilaments... e.g. to skip over and artifact. ModTwist2EM estimates fiber twist by computing the rotation of object 2 points around the axis determined by object 1 points within each contour. The overlap between adjacent contours is required to smoothly transition between contours when multiple contours are present. This duplication is not carried over to the output model, and will not bias PEET as a result. ModTwist2EM is often run immediately after addModPts, although this is not required. (If addModPts is to be used, it must be run prior to modTwist2EM).