emClarity tutorial github https://github.com/ffyr2w/emClarity-tutorial emClarity tutorial PDF https://github.com/ffyr2w/emClarity-tutorial/blob/master/emClarity-tutorial.pdf emClarity forum https://groups.google.com/g/emclarity Workflow: step00 - Tilt-series alignment, defocus estimate, subregions and picking ------------------------------------------------------------------------ Select one tilt-series, generate a first alignment, estimate the defocus and pick particles. 1) What is the "raw data"? - Look at the data directory: *.mrc, pixel size, rotation angle and tilt-scheme. - We'll select `tilt11` for the tilt-series alignment, but you can select any tilt-series you want. Note: tilt5, tilt6 and tilt8 failed the beads refinement. Note: some images need to be removed. Remove first view of tilt1-tilt10 and first and last view of tilt11 and tilt12. 2) Align the tilt-series: - e.g. ``emClarity autoAlign param_autoAlign.m ../data/tilt11.mrc ../data/tilt11.rawtlt 175 [1,41]`` - Look at the results, especially the fixedStacks directory. - 3) CTF estimate: - The tilt-series geometry has been generated, so now we can actually generate the aligned tilt-series. Also, at this point, we have enough information to estimate the (astigmatic) defocus for each view of the tilt-series. - e.g. emClarity ctf estimate param_ctf_estimate.m tilt11 - Look at alignStacks and fixedStacks/ctf 4) Subregions: - Generate the bin10: ``./recScript2.sh -1`` - Select the subregion boundaries - Generate the coordinates: ``./recScript2.sh tilt11`` - Look at ``recon/tilt11_recon.coords`` 5) Picking: - Picking in emClarity is done via template matching. You have a template ready in the directory: "EMD-5976_21.map". You can use your own template, but make sure the pixel size of that template file matches the pixel size of the aligned tilt-series.Otherwise, use ``emClarity rescale``. - Reconstruct the tomogram(s): ``emClarity ctf 3d param_templateSearch.m templateSearch`` - Look at the `cache` directory for the results. - Pick the particles: ``emClarity templateSearch param_templateSearch.m tilt11 1 EMD-5976_21.map C1 1`` - The most important results are the .csv and .mod. You can edit the .mod to remove particles using 3dmod or chimera. - Rename directory to `convmap` - (optional) Run ``emClarity init param_init.m`` step01 - Initialize project, reconstruct tomograms, subtomogram averaging and alignment --------------------------------------------------------------------------------------- You have processed one tilt-series up to the picking. To save time, we've prepared this directory, which includes the results (up to the picking) for the entire dataset. 1) Initialize the project: - ``emClarity init param_init.m`` From that point, everything you've done until this step is saved into the metadata (see `ribo.mat`). 2) Reconstruct the ctf multiplied tomograms: - ``emClarity ctf 3d param_ctf_3d.m`` Look at the `cache` directory. 3) First reconstruction and alignment cycle: - ``emClarity avg param_avg_ali_00.m 0 RawAlignment`` Check the map, FSC folder. Mention the ``emClarity fsc`` command. - ``emClarity alignRaw param_avg_ali_00.m 0`` Check alignResume to introduce the reverse order or divide option. 4) Start a new cycle: - ``emClarity avg param_avg_ali_01.m 1 RawAlignment`` step02 - Tilt-series geometry refinement ---------------------------------------- You went through an entire cycle of averaging-alignment. This directory includes results up to cycle 5. At this point the alignment at bin5 is "good enough" and we can try to refine the tilt-series geometry. 1) (optional) Remove duplicate. 2) ``emClarity tomoCPR param_avg_ali_05.m 5``. Look at mapBack1. 3) ``emClarity ctf update param_avg_ali_05.m``. Look at aliStacks and fixedStacks/ctf for *ali2*. 4) Then we need to reconstruct the tomograms with the new geometry. ``emClarity ctf 3d param_avg_ali_06.m`` You can then start cycle 6, as we've seen before. step03 - Classification ----------------------- We'll go through that section together. step04 - Final reconstruction ----------------------------- 1) At this point, the alignment had converged, so we can run FinalAlignment to generate the final map. This took 2 hours and since the command is similar to a classic RawAlignment, this directory already contains the results. 2) Run ``emClarity reconstruct param_cisTEM_recons.m 42 final_ribo C1 30`` for a "pseudo-subtomogram reconstruction" (10min).