This is Jonathan Vandezande, the Director of Computational Chemistry here at Rowan. Today I will be providing an introduction to using the Multistage Optimization workflow on the Rowan platform. Optimizations in computational chemistry can be very expensive, and it is often recommended to use a different method for optimization and single points to reduce the computational cost or increase the accuracy of computations. Rowan's Multistage Optimization workflow provides Pareto optimal combinations of methods that span the fast to accurate frontier. We spent a lot of time choosing the best modes for a wide variety of systems, preventing users from having to memorize a bunch of magic methods. These modes cover all of the first 86 elements in the periodic table and are excellent for computing the geometries and energies of molecules, non-covalent interactions, and transition states. To run a Multistage Optimization, log into the Rowan platform at labs.rowansci.com and select new Multistage Optimization. Enter the molecule or system you are interested in with one of our input methods. In this tutorial we will upload a butane molecule via SMILES. Here we have the trans configuration of butane, so we will change the name to "Butane - trans". We can check that the charge and multiplicity are correct, and then choose the type of method we would like to do. We can choose to run a transition state as the multistage optimization supports both traditional and transition state optimizations and compute frequencies. We can also choose our mode. The Rowan platform supports four different optimized modes. These are typically RECKLESS, RAPID, CAREFUL, AND METICULOUS. The first two are good for exploratory work and generating initial data. Careful is useful for publication quality results and meticulous is only used for the most demanding of situations. We can see the methods that make up these modes by mousing over the info box. RECKLESS utilizes a GFN-FF optimization with GFN2-xTB single points, generating very quick results that can direct the optimization of your work. RAPID uses a GFN2-xTB optimization with an r²SCAN-3c single point, combining the optimization speed of a semi-empirical method with the accuracy of DFT single points. This is the mode that I use the most often in my own research and strikes an excellent balance between speed and accuracy. CAREFUL runs a DFT optimization with r²SCAN-3c and a single point with ωB97X-3c and produces excellent publication quality results. METICULOUS stacks multiple DFT optimizations to quickly converge on the minimum without requiring a costly optimization with the highest accuracy method. This mode is typically only used when the utmost accuracy is needed and can be very expensive for larger systems. For this demonstration we will use the RECKLESS mode. I will go ahead and submit this job now and we can see it start running. Since this is running in the cloud on compute servers managed by Rowan, the job will run very quickly. We can see here our molecule has already optimized at the GFN-FF level of theory and performed a GFN2-xTB single point. We have our total electronic energy listed here and if we had run frequencies we would also be able to visualize them. We can check the charge tab to see the Mulliken charges of our molecule. Here we can of course see that our carbons are slightly negative and our hydrogens are slightly positive. We can then run a resubmission to compare this to the gauche conformation. Here I click the resubmit and resubmit as multistage optimization and change the name to "Butane - Gauche" We can then edit the molecule and select edit bond angle or dihedral to change the dihedral angle of this carbon backbone. Here we can drag to turn this into the gauche conformation, putting it in the approximately right condition and then selecting optimize with GFN-FF to further get a good starting guess geometry. We can then click save and then submit MultiStage Workflow. As this is running in the cloud again we can see now here our gauche conformation of the molecule. In a few seconds it should be fully optimized and we can see the fully optimized energy here. We can then select both of them to see the relative energies here with the gauche conformation being about 0.6 kcal/mole higher in energy, as would be expected. We can overlay the structures to see the differences as one of them here is trans and one of them is gauche.