Running Macroscopic pK_a Predictions in Rowan

Rowan's macroscopic pK_a workflow uses Starling, a physics-informed neural network, to compute macroscopic pK_a values and microstate populations, in addition to predicting isoelectric points, logD/pH profiles, pH-dependent aqueous solubility, and blood–brain-barrier (BBB) permeability. (For a more in-depth explanation of how this workflow works and why it matters, see our science page and our blog post about the difference between microscopic and macroscopic pK_a.)

To create a new macroscopic pK_a workflow, find this screen in Rowan:

To start submitting the workflow, click "Macroscopic pK_a, logD, and BBB prediction."

Adding a molecule

The macroscopic pK_a workflow only takes in individual 2D structures. You can add these by inputting SMILES strings, using our 2D editor, or by adding a 3D molecule through a file or the editor then clicking the "Convert to SMILES" icon to the right of the charge and multiplicity selectors.

Configuring the workflow

Charge and pH

You can configure the pH range the macroscopic pK_a workflow will analyze with the "Min pH" and "Max pH" settings.

Similarly, you can configure the minimum and maximum charges that microstates might have using the "Min charge" and "Max charge" settings.

Optional outputs

The macroscopic pK_a workflow can, in addition to predicting pK_a and logD, calculate pH-dependent aqueous solubility, solvation energy, and blood–brain-barrier (BBB) permeability (you can learn more about BBB permeability in this video). To enable pH-dependent aqueous solubility or solvation energy and BBB permeability, you can click on either of their corresponding switches.

Submitting the workflow

When your molecules are prepared and the workflow is properly configured, click "Submit macro pK_a" to submit the workflow.