TutorialSeaCoast2024

Content:

• Introduction
• Example data
• Running autoPROC (process)
• Looking at results (summary.html)
• Running with your own data

---

Introduction

Please see also:

• Looking at autoPROC results from SynchWeb (ISPyB) at Diamond Light source (DLS)

When you log into the workshop computers and open a terminal, our software should be already set up and working - just test this via.

    process -h

which should give you the command-line help for autoPROC (you might want to widen your terminal window to get nicely formatted output).

---

Example data

If you don't have your own data (old or newly collected) to play with, you can also use the example data provided with the MOSFLM and Dials tutorials

Running autoPROC (process)

You can run these via

    process -I ~/seacoast/tutorials/mosflm -d process_ex1.01 | tee process_ex1.01.lis

or

    process -I ~/seacoast/tutorials/dials -d process_ex2.01 | tee process_ex2.01.lis

What did we do here? We started a command-line program ("process") and gave it some command-line argiments (where the argument starts with a "-" (dash/minus) and a value for this argument is given afterwards (space-separated). The two arguments we used were

• -I followed by the directory name where our images are

• here we used the shell shortcut ~ (tilde) to point back to our home directory
• you could also have used $HOME/seacoast/tutorials/dials for the same effect
• or if you are already sitting in the home directory then using seacoast/tutorials/dials would also point to the same directory (as a so-called relative path)

• -d followed by a string that will be used to create a directory where all the output will go

• it is good to give it some meaningful (for you) name (here: process_ex2
• maybe use some numbering system in case you want to run the processing several times (fine-tuning some options)
• a 2-digit numbering with leading zeros has the advantage of allowing up to 99 runs while still sorting nicely in a directory listing (ls -l).

• we also wanted to save all the information printed to the terminal (so-called "standard output")
• this could be done via the "> file" syntax - but this would have the disadvantage of us not seeing what is happening while the program is running
• so we are using a pipe ("|") to hand standard output onto another program (tee) that acts like a T-junction, i.e. writes to the terminal and at the same time saves it to the file given as an argument (here process_ex.01.lis)
• make sure that output directory name (-d argument) and saved standard output are related to avoid confusion years later when you need to write up your methods sections and/or need to deposit your data to the PDB

Looking at results (summary.html)

autoPROC will give a lot of useful information on standard output - but this is "just" text and can be confusing to newcomers.

A better way of looking at the decisions and analysis provided by autoPROC is provided via a HTML file (summary.html) within the output directory given with the -d flag. This will be reported on standard output (and you should be able to just right-mouse-click onto that string to "Open link"), but it is also easy to just open a browser and set the full path of that file in the URL/location bar - here this would be

~/process_ex1.01/summary.html

or

~/process_ex2.02/summary.html

(if you did run this in your home directory).

---

Running with your own data

First decide where you want to have results - some kind of organised directory structure can be useful here, e.g.

    mkdir -p ~/Projects/Lysozyme/DLS-Data/20240131
    cd ~/Projects/Lysozyme/DLS-Data/20240131

(this is just an example, organised by protein, synchrotron and data collection date as YYYYMMDD ... whatever system you choose: make it meaningful and then stick with it).

You just need to point to the directory with your images (coming from a single crystal) and define output directory and logfile:

    process -I ~/Images/DLS-Data/20240131 -d process.01 | tee process.01.lis

(of course the exact path to the image directory is different for you).

Then open the symmary.html file via e.g.

    firefox ~/Projects/Lysozyme/DLS-Data/20240131/process.01/summary.html

or (easier): open a new terminal tab (sould place you into the same directory) and hten run

    firefox process.01/summary.html