Introduction

Acid dissociation constant (pKa) values indicate the degree to which a molecule will be ionized (charged) or neutral at a given pH, such as the physiological pH (~7.4). This ionization state directly impacts numerous crucial properties of a drug candidate, including solubility, membrane permeability, target binding affinity, and ADMET properties.

The Small Molecule pKa Prediction Floes predict the pKa of small molecules and serve different needs in regard to those predictions. The underlying prediction is based on the Gaussian Process Regression model trained on a curated proprietary dataset.

There are three floes in this package.

1. Generate Ionization States and Calculate LogD Floe

   This is a dataset preparation floe. It predicts micro- and macro-pKas, and based on those values, also calculates the most dominant states of the molecule and LogD at a chosen pH. This floe is suitable to run on a few tens of thousands of molecules.

2. pKa Exploration Floe

   This floe predicts pKa values and generates a detailed Floe Report. It is intended to run on a few hundreds of molecules since it generates large reports.

3. Prepare Supporting Experimental pKa Data File Floe

   This floe allows users to utilize their own experimental data to affect or improve predictions with the model employed in this package.