Some additional notes:
Job001 import is for generating a movie list of eer movies. These were subsequently converted to tiff using relion_convert_to_tiff. Will save you 2/3 data space. Job002 are the converted micrographs. 

Initially a threshold of -7 was trialled as a too low threshold resulted in less filament B type being picked. This resulted in a large number of particles which meant refinements took very long. A threshold of -5 was also trialled. This showed that there were still filament type B picked. As such, -5 is the better option, less junk is picked. Although both approaches would be okay, this shows that picking less is better and will save you time.

#Commands for initial model generation of filament A (the triplet)
#rhi is an alias for relion_helix_inimodel and ri is for relion_image_handler 
rhi --i Select/job051/class_averages.star --o inimodel/testA --angpix 4.944 --mask_diameter 300 --j 2 --crossover_distance 700
xmipp_show -img before_reproject.spi after_reproject.spi rec.spi -poll &
rhi --i Select/job054/class_averages.star --o inimodel/testA --angpix 4.944 --mask_diameter 300 --j 2 --crossover_distance 700
rhi --i Select/job054/class_averages.star --o inimodel/testA-C3 --angpix 4.944 --mask_diameter 300 --j 2 --crossover_distance 750 --iniref inimodel/testA_it010_reconstructed.mrcs --sym 3
#the reconstructed map needs to be rescaled for refinement 
ri --i inimodel/testA-C3_class001_rec3d.mrc --o ini4refine/testA-C3-750CO.mrc --rescale_angpix 0.824 --new_box 384