The SAMPL7 pKa Challenge consists of predicting microscopic pKa's of small molecules. Participants are asked to report relative free energies of protonation microstates calculated based on a provided reference state and pH detailed below. Relative free energies of microstates were chosen as the reporting format rather than the microscopic pKa values given the recent work of Gunner et al.[1] The submission deadline is Oct. 8, 2020.
For each molecule, the relative free energy must be predicted between the reference microstate and the rest of the enumerated microstates for that molecule at a reference pH of 0. In this case, we are after the relative state free energy including the proton free energy (if that is helpful in clarifying), which could also be called the reaction free energy for the microstate transition which has the reference state (SMXX_micro000) as the reactant and the alternate state as one of the products. If
The first microstate in each CSV file indicated by SMXX_micro000 is our selected neutral reference state. For example, for molecule SM25, if the reference microstate is SM25_micro000, then relative free energies must be computed between SM25_micro000 and SM25_micro001, SM25_micro000 and SM25_micro002, and SM25_micro000 and SM25_micro003 (at a pH of 0). For compounds SM35, SM36 and SM37 the issue of reference state becomes more complex because of chirality. Our chosen reference states had an unspecified stereocenter for these compounds, and we updated this repo on October 1, 2020, to make clear that we intended the enantiopure form of these compounds. As such, you may assume that the state with unspecified stereochemistry is at the same free energy as the microstate with specified stereochemistry, e.g. that SM35_micro000 is equal in free energy to the correct state SM35_micro002.
All possible tautomers of each ionization (charge) state are defined as distinct protonation microstates.
If you are following the work of Gunner et al. to understand this format, please note that the study contains some sign errors for ΔG values, and indexing could perhaps use some clarification. In the notation of that paper, ΔGij is the free energy for the transition where state j is the reference state, e.g. the free energy for the reaction where j is a reactant and i is a product. Here we are asking for that same reaction free energy, where state j is SMXX_micro000 and the other states you consider are compared relative to this state.
On Sept. 30, 2020, we updated this repo to add some additional potential microstates identified by Bogdan Iorga. You are at liberty to include or neglect these microstates in your submission. However, please note that any microstates neglected in your submissions will be assumed to have negligible population relative to those included -- i.e. we will assume that the free energy for transition to those microstates is large and unfavorable.
We would like to note that compounds SM35, SM36 and SM37 are enantiopure and have a chiral center. All other compounds are not chiral. The files provided in this repository provided versions of these compounds with both specified and unspecified stereochemistry for these compounds because we were not clear that they had been tested in enantiopure form. This was corrected on Oct. 1, 2020. We retain the full set of files here for historical reasons, but predictions should use the files with specified chirality. Note, as above, you may assume that microstates with unspecified chirality are equal in free energy to those with specified chirality.
A submission template file can be found in the submission_template/ directory and an example submission file can be found in example_submission_file/. Predictions must be submitted via our AWS submissions server, http://sampl-submit.us-west-1.elasticbeanstalk.com/submit/SAMPL7-pKa. Note that as of Sept. 28, 2020, this format is being corrected/changed/clarified. Please follow our email list (be sure to indicate interest in SAMPL7 and pKa) to get a notification when we fix the format and update instructions, likely Sept. 30 or Oct. 1.
-
Fill one
submission_template/pKa_prediction_template.csvtemplate for all molecules predicted with one method. You may submit predictions from multiple methods, but you should fill a separate template file for each different method. -
Record the pair of microstates IDs associated with each relative free energy calculation between the reference state and the predicted microstates. Enumerated microstates, IDs, SMILES strings and SDF/MOL2 files can be found in https://github.com/samplchallenges/SAMPL7/tree/master/physical_property/pKa/microstates.
-
We highly recommend including your predicted macroscopic pKa values for each compound in your
Methodsection in the specified format; see template. (We added this suggestion Sept. 30, as it will allow us to check that our analysis arrives at the same macro pKa values as yours does.) -
If you have evaluated additional microstates, we ask that you include the same information as the other challenge molecules, but include the SMILES string in your submission and email a
.mol2file of the microstate with explicit hydrogens and correct bond orders to[email protected]. Additonal microstate molecule ID's must be in the formSMXX_extra001, where the molecule tagSMXXis followed by_extraand some number001with three characters. Seeherefor an example. -
You may report only 1 relative free energy value per molecule per method.
-
Each participant or organization is allowed only one ranked submission.
-
Anonymous participation is not allowed.
-
The energy units must be in kcal/mol.
-
It is mandatory to submit predictions for all 22 molecules. Incomplete submissions will not be accepted.
-
Report relative free energy values to two decimal places (e.g. 13.71).
-
Report the standard error of the mean (SEM) as a measure of statistical uncertainty (imprecision) for your method. The SEM should capture variation of predicted values of the same method over repeated calculations.
-
Report the model uncertainty of your difference in free energy prediction --- the predicted accuracy of your method [2,3]. This is not a statistical uncertainty. Rather, the model uncertainty is an estimate of how well your predicted values are expected to agree with experimental values. For example, for classical simulation approaches based on force fields, this could measure how well you expect the force field will agree with experiment for this compound. The model uncertainty could be global or different for each molecule. For example, reference calculations in SAMPL5 log D challenge estimated the model uncertainty as the root mean squared error (RMSE) between predicted and experimental values for a set of molecules with published cyclohexane-water partition coefficients.
-
Lines beginning with a hash-tag (#) may be included as comments. These and blank lines will be ignored during analysis.
-
The file must contain the following four components in the following order: your predictions, a name for your computational protocol (that is 40 characters or less), the average compute time across all of the molecules in hours (GPU/CPU time for physical methods, query time for empirical methods), details of the computing resources and hardware used to make predictions, a list of the major software packages used, prediction method category, and a long-form methods description. Each of these components must begin with a line containing only the corresponding keyword:
Predictions:,Participant name:,Participant organization:,Name:,Compute time:,Computing and hardware:,Software:,Category:,Method:, andRanked:, as illustrated in the example file. An example submission file can be found here to illustrate expected format when filling submission templates. -
For Method Category section please state if your prediction method can be better classified as an experimental database lookup (DL), linear free energy relationship (LFER)[3], quantitative structure-property relationship or machine learning (QSPR/ML)[3], quantum mechanics without empirical correction (QM) models, quantum mechanics with linear empirical correction (QM+LEC), and combined quantum mechanics and molecular mechanics (QM+MM), or Other, using the category labels
DL,LFER,QSPR/ML,QM,QM+LEC,QM+MMorOther. If you choose the “Other” category, please explain your decision in the beginning of Method Description section. -
Names of the prediction files must have three sections separated by a
-: predicted propertypKa, and your name and must end with an integer indicating the number of prediction set. For example, if you want to submit one prediction, you would name itpKa-myname-1.csv, wheremynameis arbitrary text of your choice. If you submit three prediction files, you would name thempKa-myname-1.csv,pKa-myname-2.csv, andpKa-myname-3.csv. -
Prediction files will be machine parsed, so correct formatting is essential. Files with the wrong format will not be accepted.
As per our policy on multiple submissions, each participant or organization is allowed only one ranked submission, which must be clearly indicated as such by filling the appropriate field in the submission form. We also accept non-ranked submissions, which we will not formally judge. These allow us to certify that your calculations were done without knowing the answers, but do not receive formal ranking, as discussed at the link above.
If multiple submissions are incorrectly provided as "ranked" by a single participant, we will judge only one of them; likely this will be the first submitted, but it may be a random submission.
pKa measurements were obtained via automated potentiometric titrations using a Sirius T3 instrument (Pion, Inc)[5] by the Ballatore lab at UCSD. Three titrations were performed from pH 1.8 to pH 12.2 using ionic strength adjusted water (0.15 M KCl), acid (0.5 M HCl, 0.15 M KCl) and base (0.5 M KOH, 0.15 M KCl). The pKas of select compounds (Compounds SM30 and SM39) with low aqueous solubility were measured using a cosolvent protocol; Yasuda-Shedlovsky extrapolation method was used to estimate the pKa at 0% cosolvent.
Experiments using the Sirius T3 were done at 25°C. pKa determination with the Sirius T3 were done using solid samples, not in solutions of specific concentration, so the concentrations were different per compound. pKa's were determined from 3 or more titrations.
Your method descriptions should give a detailed description of your approach, ideally with enough detail that someone could reproduce the work. These often serve to allow researchers to coordinate on why calculations which seem similar performed quite different in practice, so you should be sure to address how you generated poses, selected protonation states and tautomers if applicable, dealt with counterions, and various other aspects that might be important, as well as any method-specific details that, if varied, might result in different performance. For example, with MD simulations, the amount of equilibration might impact performance significantly in some cases, so this should also be included.
As per above, we are also highly recommending you include your predicted macro pKa values in this section in the specified format (see template) to allow us to use these for consistency checking later.
You may use any method(s) you like to generate your predictions; e.g., molecular mechanics or quantum mechanics based methods, QSPR, empirical pKa prediction tools etc.
Some participants use SAMPL to help evaluate various computational methods. To accommodate this, multiple prediction sets from a single research group or company are allowed, even for the same type of predictions if they are made by different methods. If you would like to submit predictions from multiple methods, you should fill a separate submission template files for each different method.
[1] Gunner, M.R., Murakami, T., Rustenburg, A.S. et al. "Standard state free energies, not pKas, are ideal for describing small molecule protonation and tautomeric states." Journal of Computer-Aided Molecular Design 34, 561–573 (2020). https://doi.org/10.1007/s10822-020-00280-7
[2] Bannan, Caitlin C., Kalistyn H. Burley, Michael Chiu, Michael R. Shirts, Michael K. Gilson, and David L. Mobley. “Blind Prediction of Cyclohexane–water Distribution Coefficients from the SAMPL5 Challenge.” Journal of Computer-Aided Molecular Design 30, no. 11 (November 2016): 927–44.
[3] Mobley, David L., Karisa L. Wymer, Nathan M. Lim, and J. Peter Guthrie. “Blind Prediction of Solvation Free Energies from the SAMPL4 Challenge.” Journal of Computer-Aided Molecular Design 28, no. 3 (March 2014): 135–50. https://doi.org/10.1007/s10822-014-9718-2
[4] U. A. Chaudry and P. L. A. Popelier. “Estimation of pKa Using Quantum Topological Molecular Similarity Descriptors: Application to Carboxylic Acids, Anilines and Phenols.” The Journal of Organic Chemistry 2004 69 (2), 233-241. https://doi.org/10.1021/jo0347415
[5] Işık, M., Levorse, D., Rustenburg, A.S. et al. "pKa measurements for the SAMPL6 prediction challenge for a set of kinase inhibitor-like fragments." Journal of Computer-Aided Molecular Design 32, 1117–1138 (2018). https://doi.org/10.1007/s10822-018-0168-0