Studies of the relative stability of TFA adducts vs non-TFA analogues for combinatorial chemistry library members in DMSO in a repository compound collection

Maintaining a High-Quality Screening Collection: The GSK Experience

The storage of screening collections in DMSO is commonplace in the pharmaceutical industry. To ensure a high-quality screening collection, and hence effective and efficient high-throughput screening, all compounds entering the GlaxoSmithKline (GSK) screening collection undergo a liquid chromatography-mass spectrometry (LC-MS) quality control (QC). It is generally accepted that even under optimal conditions, a small percentage of these compounds are unstable after prolonged storage in DMSO. This article presents how these QC data can be mined using a data-driven clustering algorithm to identify chemical substructures likely to cause degradation in DMSO. This knowledge provides new structural filters for use in excluding compounds with these undesirable substructures from the collection. This information also suggests an efficient, targeted approach to compound collection clean-up initiatives. Stability studies are also designed to maintain a high-quality screening collection. To define the best practice for the storage and handling of solution samples, GSK has undertaken stability experiments for two decades, initially to support the implementation of new automated liquid stores and, subsequently, to enhance storage and use of compounds in solution through an understanding of compound degradation under storage and assay conditions.

A Novel Approach to Determine Water Content in DMSO for a Compound Collection Repository

Mail Stop 29-M-B Thousand Oaks, CA 91320 The quality of a corporate compound collection can be significantly affected by a complex combination of storage and operational processing factors. Water content in DMSO solutions is one factor that is of great interest as it can affect solubility, degradation, and freeze-thaw cycle parameters. To the authors’ knowledge, this is the first report of using near-infrared (NIR) spectroscopy to assess water content in DMSO compound stock solutions within the common storage vessel format of polypropylene microtubes. The precision and accuracy of the NIR technique was benchmarked against a Karl Fisher titration method, and a correlation coefficient was determined to be 0.985 over a range of 1% to 10% water in DMSO by weight. The advantages of the NIR technique include accuracy, precision, speed, nondestructiveness, and the capability of assessing compounds under in situ storage conditions within microtubes. In this report, the authors demonstrate the accuracy and precision of using NIR to assess water content in DMSO solutions and present a case study to demonstrate the utility of the technique to aid in assessing a pharmaceutical compound collection.

Computational prediction and validation of an expert's evaluation of chemical probes

In a decade with over half a billion dollars of investment, more than 300 chemical probes have been identified to have biological activity through NIH funded screening efforts. We have collected the evaluations of an experienced medicinal chemist on the likely chemistry quality of these probes based on a number of criteria including literature related to the probe and potential chemical reactivity. Over 20% of these probes were found to be undesirable. Analysis of the molecular properties of these compounds scored as desirable suggested higher pKa, molecular weight, heavy atom count, and rotatable bond number. We were particularly interested whether the human evaluation aspect of medicinal chemistry due diligence could be computationally predicted. We used a process of sequential Bayesian model building and iterative testing as we included additional probes. Following external validation of these methods and comparing different machine learning methods, we identified Bayesian models with accuracy comparable to other measures of drug-likeness and filtering rules created to date.

ChemStable: a web server for rule-embedded naïve Bayesian learning approach to predict compound stability

Predicting compound chemical stability is important because unstable compounds can lead to either false positive or to false negative conclusions in bioassays. Experimental data (COMDECOM) measured from DMSO/H2O solutions stored at 50 °C for 105 days were used to predicted stability by applying rule-embedded naïve Bayesian learning, based upon atom center fragment (ACF) features. To build the naïve Bayesian classifier, we derived ACF features from 9,746 compounds in the COMDECOM dataset. By recursively applying naïve Bayesian learning from the data set, each ACF is assigned with an expected stable probability (p(s)) and an unstable probability (p(uns)). 13,340 ACFs, together with their p(s) and p(uns) data, were stored in a knowledge base for use by the Bayesian classifier. For a given compound, its ACFs were derived from its structure connection table with the same protocol used to drive ACFs from the training data. Then, the Bayesian classifier assigned p(s) and p(uns) values to the compound ACFs by a structural pattern recognition algorithm, which was implemented in-house. Compound instability is calculated, with Bayes' theorem, based upon the p(s) and p(uns) values of the compound ACFs. We were able to achieve performance with an AUC value of 84% and a tenfold cross validation accuracy of 76.5%. To reduce false negatives, a rule-based approach has been embedded in the classifier. The rule-based module allows the program to improve its predictivity by expanding its compound instability knowledge base, thus further reducing the possibility of false negatives. To our knowledge, this is the first in silico prediction service for the prediction of the stabilities of organic compounds.

Early phase drug discovery: cheminformatics and computational techniques in identifying lead series

Early drug discovery processes rely on hit finding procedures followed by extensive experimental confirmation in order to select high priority hit series which then undergo further scrutiny in hit-to-lead studies. The experimental cost and the risk associated with poor selection of lead series can be greatly reduced by the use of many different computational and cheminformatic techniques to sort and prioritize compounds. We describe the steps in typical hit identification and hit-to-lead programs and then describe how cheminformatic analysis assists this process. In particular, scaffold analysis, clustering and property calculations assist in the design of high-throughput screening libraries, the early analysis of hits and then organizing compounds into series for their progression from hits to leads. Additionally, these computational tools can be used in virtual screening to design hit-finding libraries and as procedures to help with early SAR exploration.

On-column solvent exchange for purified preparative fractions

On-column solvent exchange, using many of the principles of solid-phase extraction, has been implemented to significantly reduce evaporation cycle time following reverse-phase preparative HPLC. Additional benefits, such as a reduced potential for salt formation, thermal decomposition, and residual solvent, are also described. Fractions obtained from preparative separations, typically in a large volume of acetonitrile:water, are injected into the preparative HPLC and then eluted in acetonitrile, creating a new fraction in a volatile organic solvent. Minimal modification to the instrument was required, and unattended operation is possible. Acetonitrile evaporation is achieved within 3 h, compared with 17 h for aqueous-based fractions; lower temperatures can be used during the evaporation step; mobile-phase additives, likely to form salts with the target compound if concentrated in the fraction, are removed before evaporation; sample recovery and purity are unaffected.

Monitoring compound integrity with cytochrome P450 assays and qHTS

The authors describe how room temperature storage of a 1120-member compound library prepared in either DMSO or in a hydrated-DMSO/water (67/33) mixture affects the reproducibility of potency values as monitored using cytochrome P450 1A2 and 2D6 isozyme assays. The bioluminescent assays showed Z' factors of 0.71 and 0.62, with 17% and 32% of the library found as active against the CYP 1A2 and 2D6 isozymes, respectively. The authors tested the library using quantitative high-throughput screening to generate potency values for every library member, which was measured at 7 time intervals spanning 37 weeks. They calculated the minimum significant ratio (MSR) from these potency values at each time interval and found that for the library stored in DMSO, the CYP 1A2 and 2D6 assay MSRs progressed from approximately 2.0 to 5.0. The hydrated conditions showed similar performance in both MSR progression and analytical quality control results. Based on this study, the authors recommend that DMSO samples be stored in 1536-well plates for <4 months at room temperature. Furthermore, the study illustrates the degree and time scale of apparent compound potency changes due to sample storage.

A Kinetic Study of Product Cleavage Reactions from the Solid Phase by a Biocompatible and Removable Cleaving Reagent, HCl

TFA has been widely used as a cleaving reagent in solid-phase organic synthesis. However, it is difficult to remove from the final product, and it is toxic to various cells. To search for an alternative, we studied the kinetics of HCl cleavage reactions of 18 resin-bound compounds on various linkers. HCl is very easy to remove completely from samples, and the residual HCl does not have a toxic effect in cell assays. Most compounds studied in this work can be easily cleaved using a low concentration of HCl (0.9-2.3%) and the minimal amount of time (60-90 min). Even in the most difficult case, a moderate 8% HCl and an extended time (10-15 h) are enough to cleave the product. Therefore, our kinetic studies establish HCl as a biocompatible, removable, and effective substitute for TFA when final compounds are used for biological screening and drug discovery.

Investigation of 3 Industry-Wide Applied Storage Conditions for Compound Libraries

The appropriate storage conditions for a compound file are a crucial factor for the success of drug discovery projects. In this study, 778 highly diverse compounds dissolved in 100% DMSO were stored under 3 industry-wide accepted storage conditions, and the compound integrity was monitored for a period of 6 months. The storage conditions selected were (1) under argon at +15 °C, (2) under argon at -20 °C, and (3) under ambient atmosphere at -20 °C. Each sample was assessed every 4 weeks by liquid chromatography coupled to mass spectrometry (LC/MS). Based on the resulting experimental data, a statistical projection of compound integrity over a period of 4 years for each of the 3 storage conditions was generated applying a linear mixed-effects model. A moderate loss of compound integrity of 12% was calculated for storage at -20 °C under argon, a loss of 21% for storage at -20 °C under ambient atmosphere, and a strong decrease of 58% for storage at +15 °C under argon over this period. The initial purity of the compounds does also influence the rate of compound degradation. Compounds with an initial purity of 50% to 75% degraded faster than compounds with an initial purity of more than 75%. The results of the study enable the prediction of the point in time, when the purity of a compound population falls below a predefined threshold that would trigger the resolubilization or retirement of the compound population represented by the analyzed samples.

In silico estimation of DMSO solubility of organic compounds for bioscreening

Solubility of organic compounds in DMSO is an important issue for commercial and academic organizations handling large compound collections or performing biological screening. In particular, solubility data are critical for the optimization of storage conditions and for the selection of compounds for bioscreening compatible with the assay protocol. Solubility is largely determined by the solvation energy and the crystal disruption energy, and these molecular phenomena should be assessed in structure-solubility correlation studies. The authors summarize our long-term experimental observations and theoretical studies of physicochemical determinants of DMSO solubility of organic substances. They compiled a comprehensive reference database of proprietary data on compound solubility (55,277 compounds with good DMSO solubility and 10,223 compounds with poor DMSO solubility), calculated specific molecular descriptors (topological, electromagnetic, charge, and lipophilicity parameters), and applied an advanced machine-learning approach for training neural networks to address the solubility. Both supervised (feed-forward, back-propagated neural networks) and unsupervised (Kohonen neural networks) learning methods were used. The resulting neural network models were validated by successfully predicting DMSO solubility of compounds in independent test selections.

Comparison of Preparative HPLC/MS and Preparative SFC Techniques for the High-Throughput Purification of Compound Libraries

A diverse set of 16 high-throughput organic synthesis libraries, consisting of 48 samples per library, has been purified by both preparative supercritical fluid chromatography (SFC) and preparative high-performance liquid chromatography (HPLC). This paper details the relative effectiveness of these two purification techniques in terms of success, yield, and purity of final product.