 Hi, my name is Eli Chadwick, I'm speaking on behalf of my colleagues Leandra Laborier, Simone Sterniolo and Alejandro Gonzales Beltran on Muon Galaxy, which is our attempt to extend Galaxy for Muon Science. We work at the STFC Rutherford-Appleton Laboratory, which is home to the ISIS Neutrona Muon Source, one of only four Muon spectroscopy facilities in the world. My team and I work in the Scientific Computing Department, which supports the computational needs of ISIS and the other experiments on site, as well as developing computational tools for the wider academic community. On the right is a picture of the campus, in the front you have a few buildings that make up the ISIS experiment, and in the background is the small building that hosts scientific computing. So a brief overview of Muon spectroscopy. Muons are positively charged particles that behave essentially like small protons. You fire a beam of them at your material that you want to study, and the muons stop somewhere in the material. And then after a couple of microseconds, each Muon will decay, releasing a positron, and the positrons are what are detected by the detectors. Now where the muons stop within the material isn't random. It's dictated by the local environment within the material, so every material has quite a distinct set of stopping sites, as they're known. Now we can simulate Muon experiments, and simulations can make predictions about stopping sites, and therefore help us to predict material properties. And generally, simulations can enhance analysis of your experimental data, and enable you to use your beam time more efficiently, which is important because it's limited and expensive because there aren't many Muon facilities. However, most Muon scientists don't make use of simulations to complement their experimental work, and that's something that my team are trying to encourage. We work on the Muon Spectroscopy Computational Project, which is creating a sustainable, accessible and open source ecosystem of tools for Muon science. The main thing that we work on is called Pi Muon Suite, which is a collection of command line tools to use alongside DFT density functional theory simulations, and those model the physical properties of nanosystems. We're also now working on Muon Galaxy, which is a trial of Galaxy to interface with those Pi Muon Suite tools. And there's a link to our website here if you want more information on the project. So the Muon science pipeline has a few stages. You start out with your structure, you add a Muon to it in a random location, then you optimize to figure out where the Muon goes, and you do that maybe a hundred times. And then you look at those hundred runs altogether, and you see are there locations where the Muon tends to end up? And those are likely to be your stopping sites. And once you have that information, you can do all sorts of post processing and study material properties. So Pi Muon Suite contains command line tools for each of these stages. And what we want from a graphical user interface is to be able to run those tools individually, but also to chain them into workflows to make the whole process a bit more simple. And Galaxy seems like an ideal platform for that. So we started converting our command line tools into Galaxy tools. And this is a work in progress. So far, we've run into one significant issue, which is that our code outputs lots of nested folders, and that doesn't seem to be particularly compatible with Galaxy collections. If you have advice on this front, we'd be interested to hear it. More ambitiously, we are also developing a visualizer called ChrisViz.js, and this is for visualizing molecules. Here there's a visualizer which has loaded into it an organic molecule. I don't know exactly what it is. But the visualizer allows you to, there are options to change the view. So you can show the chemical symbols associated with each atom. Or you can click in the visualizer itself and select a particular atom. So here there's a carbon atom circled in yellow that's been selected. And that can be used to generate some output. So we're going to customize this so that the output can be fed directly into whatever web app we use for our interface. So we've tried to integrate this visualizer into Galaxy, and we've run into a couple of issues. Firstly, how to include controls in the visualizer display. I haven't looked through all the existing visualizations, so if there are any that do this, please point me to them. That would be really helpful. The second problem, which seems to be a bigger problem, is that we want an interaction where you can click an atom in the visualizer and then use that atom as input to another tool. So basically be able to say, I want to center this tool on this atom. But Galaxy doesn't support output from visualization plugins at present, so there's a question of how to actually accomplish this. Do we add output to visualization plugins? Do we embed the visualizer in a tool instead? Do we create something completely custom? We're a bit stuck here, so if you have ideas, advice, if you are interested in or doing something similar, please do get in touch. It would be really helpful to have some conversations. Thank you very much for listening. Please talk to us about running Galaxy at national facilities, importing collections from nested folders, integrated visualizer controls, and visualizes as tools or visualizes with output. My email is elai.chabwick at stfc.ac.uk if you want to get in touch. Thank you.