Frequently Asked Questions

How many molecules may I provide as input to these floes?

The Generate Ionization States and Calculate LogD Floe is intended to run on a dataset of a few tens of thousands of molecules. The pKa exploration floe is intended to run on a dataset of a few hundreds of molecules.

How are input molecules preprocessed?

Every input or query molecule is automatically preprocessed in order to begin with a neutral form. In this step, any smaller fragments and salts are removed and formal charges are removed from the largest component. Floes in this package do not process input molecules to find the more favorable tautomer.

The number of ionizing groups are assessed. The next section describes how a higher number of ionizing groups are processed. If no ionizable group is found, then the molecule is directly passed to the output unchanged, with OEXlogP reported as LogD (if the option to calculate it is selected).

What is the maximum number of ionizing groups per molecule supported by these floes?

Since the number of microstates (\(2^n\)) and microtransitions (\(n \cdot 2^{n-1}\)) increases drastically with the number of ionizable groups, there are limitations to the number of ionizable groups that can be processed in detail when generating all possible microstates and microtransitions.

The default cutoff for the ionizable groups count (n) is 4. It can be used in the following way. All floes in this package first quickly estimate the intrinsic pKa for each ionizable group. Intrinsic pKa is the pKa of one deprotonating group while all other groups are kept in an uncharged form. After assessing all intrinsic pKa values, the floes filter out groups with pKa < 0.0 and > 14.0. If the number of remaining groups (0.0 <= intrinsic pKa <= 14.0) are still more than the cutoff for the ionizable groups count n, then n groups which are closest to the set environment pH (default 7.4) are chosen to perform detailed microstate and microtransition analysis.

The unselected (filtered out) groups are still reported in exploration Floe Reports, and their favored ionization states (as per intrinsic pKa and environment pH) are adjusted for all microstates. The Cutoff for Ionizable Groups Count parameter can be modified in the Advanced Options parameters in both the Generate Ionization States and Calculate LogD and pKa Exploration Floes.

Could you show an example using an unselected ionizing group for pKa prediction?

This example shows protonation microequilibria that are generated with an unselected ionizing group for detailed micro- and macro-pKa prediction. Figure 1 shows a molecule where a few ionizing groups were excluded from detailed micro- and macro-pKa calculations.

Figure 1. A protonation microequilibria example with an unselected ionizing group for detailed micro- and macro-pKa prediction.

The ionizing groups with an intrinsic pKa outside the range of 0 to 14, or which were filtered out due to too many functional groups, are shown with their intrinsic pKa as micro- and macro-pKa values. All remaining selected groups for detailed micro- and macro-pKa prediction are shown with all possible microtransitions for those groups from all protonated to all deprotonated.

Which floe should be used to predict pKa values without adjusting ionization states?

To calculate only the pKa (assumed to be macro-pKa) values and not generate a dominant state or detailed reports, one should run the Generate Ionization States and Calculate LogD Floe and save the significant state in the new column by providing a name in the Significant Ionization State Molecule Field output parameter as shown in the tutorial for that floe.

Which floe should be used to predict the LogD value at a chosen environmental pH without adjusting the ionization states?

To calculate only LogD, follow the same directions given in the question above.