#######################################
workshop instruction, for details go to
#######################################
Red command lines are matlab commands.
Blue command lines are linux commands.
Installation of Dynamo_1.1.288 for matlab:
tar xvf dynamo-v-1.1.288_MCR-9.2.0_GLNXA64_withMCR.tar
you'll need matlab 2017a, cuda 7.5.
We will use dynamo GUI from matlab for particle picking and setting up parameters, and will run dynamo iterations in matlab with GPUs.
If you open dynamo-1.1.288, there is a folder called README_dynamo_installation.txt. Follow the instructions on these points:
1) MATLAB (all platforms: Linux and Mac)
2A) LINUX (standalone)
4) GPU
5) MCR libraries
When you run commands of 2), 4), and 5), please go to bash shell.
First activate dynamo in matlab:
run ~/dynamo_directory/dynamo_activate.m
################## catalogue creation and model building #####################
Use the volume list file to build a catalogue for tula tomogram data. First change “tomograms_directory” in the vll file to the correct directory of the tomograms, then
dcm -c tula_workshop_140814 -vll tula_workshop_140814.vll
Open dynamo catalogue to work on the data: pick membrane using either vesicle or ellipsoid vesicle model
dcm -c tula_workshop_140814
organize data using dynamo catalogue, then pick membrane using either vesicle or ellipsoid vesicle model
first select tula_140814_tomo01_bin2_lp60.mrc > view volume preview [on-the-fly load]
use contrast panel to optimize contrast
to save a region:
use load fragment panel to select regions > regions > save current fragment as region in catalogue > save under a name like 'tomo0101'
to load the region into picking tool:
load (memory)dynamo: tomoview > change thickness to 1 in the viewing direction panel > change depth to 8 in the scene panel > optimize contrast in necessary
to start picking:
model pool > create new model in pool (choose type) > choose vesicle or ellipsoidal vesicle depending on the virus geometry > go to different slices on viewing direction panel, place the cursor on where you want to pick, and press c on the keyboard to mark, for vesicle model, do this for 3-5 slices on the virus, cover as far as you can still see the virus > press v on the keyboard if you want to zoom in the place.
To edit model:
active model edit active model > press the first button to calculate centre and radius of the model > press the 2nd button to divide points > change mesh parameter (depiction) to 10 to reduce mesh numbers > press the 6th button to create mesh > plot mesh to see the mesh created > press the 9th button to refine mesh > plot mesh to see the mesh created > press the 10th button to create cropping mesh (change mesh parameter (cropping)) to adjust the mesh number to ~200-300 > plot cropmesh to see the mesh created > press the 11th button to uniformaize the cropping mesh > plot cropmesh to see the mesh created > press the 14th button to create vectors > plot cropmesh and tablesketch to see the mesh created > edit save table on the data particles panel to save the table file > on tmview active model save model into catalogue (disk) > on model pool clear current model pool from memory > close the tmview, active model editors and work on another region.
To crop particles:
on catalogue manager select all labe > crop particles > open volume list manager > choose all models > edit sidelength to 80 and edit filename to tula_workshop_140814_crop.vll > create list > open the list and erase _lp60 from the name
dtcrop tula_workshop_140814_crop.vll reorder tula_workshop_140814_seeds 80 -ws o
To create an initial average as the template for STA:
dynamo_average('tula_workshop_140814_seeds','table','./ tula_workshop_140814_crop.tbl','fcompensate','1','o','tula_stage1_ite0.em','actions','nmask');
######################### project building ###########################
In matlab:
dcp
Feed particles, table, template, masks and parameters into the panel.
######################### running a project on GPUs ###########################
To run project on matlab GPUs with dynamo 1.1.288, first open a terminal on a computer with GPU cards
source /cuda_root/cuda.cshrc
then open matlab 2017a and activate dynamo
run /home/user/dynamo-v-1.1.288_MCR-9.2.0_GLNXA64_withMCR/dynamo_activate.m
cd /home/user/Desktop/IWAIP2017/datasets_IWAIP2017/tomograms_tula_testdata_IWAIP2017_SaiLi
dvput workshop_tula1 -destination matlab_gpu -gpu_identifier_set [0,1,2]
dvcheck workshop_tula1
dvunfold workshop_tula1
run workshop_tula1
######################### Checking results ###########################
To check the averages:
dmapview average_ref_001_ite_0002.em
To check the tables:
dtview -t refined_table_ref_001_ite_0006.tbl
To plot the templates onto the coordinates:
dtplot table.tbl -m chimera –tr reg=1 -template template.em
######################### Sub-boxing ###########################
To preview the subbox points:
dynamo_sym_point [29,40,38] c4 -template average_lp20p.em -show template;
To do subboxing:
ddsubboxing /data_directory 36 -r [29,40,38] -t refined_table_ref_001_ite_0001.tbl -st 80 -sym c4 -rsr [90,-7,-90] -o tula_subbox
dynamo_average('data','table','./subbox_table.tbl','fcompensate','1','o','template.em','actions','nmask');
Useful functions:
dynamo_resolution('apix',5.4,'k',8,'n',80)