What is High-Throughput Purification?

Introduction

High-throughput purification (HTP) is a set of separate workflow stages that combine to rapidly and efficiently purify a large number of potential drug candidates from libraries of thousands of compounds.  The development of HTP has been a major advancement in drug discovery over the last five to ten years, significantly accelerating the drug discovery process and leading to the development of new drugs that would not have been possible otherwise.

The goals in HTP are often slightly different than those of singleton purification: why might this be? Typically for singleton purifications, each sample being purified can be treated as a unique purification with optimization at each stage of the process. This process is time- and personnel- intensive, however, it does offer the highest possible rate of success with the highest purity of the final purified materials.  The business case for HTP is to purify as many samples possible as quickly as possible, driven by the throughput of large numbers of samples (100's to 1,000's) which are often in microtiter plates.  

The number of samples purified to a high purity is governed by how much upfront screening is done to select the separation methods that best isolate the compound of interest from any closely eluting impurities. Depending on the diversity of chemistries being purified, scientists may screen crude samples using a single chromatographic method that is known to give a good general separation coverage. Others might screen under acidic and basic conditions and in some cases a mixture of (U)HPLC and SFC methodologies are combined to explore the potential for a more complete isolation of the compound of interest.

Software Automation in High-Throughput Purification

Implementation of automated software-driven interpretation and decision making in HTP can increase throughput without increasing head count or reducing the quality final registered product.

The higher the number of methods that are screened, the greater the chance of being able to identify a separation method that returns the highest purity and recovery. More injections per sample of course, also means more data to process and evaluate and more time to acquire the sample data in the first place.

This is where automated software-based interpretation and decision making can alleviate the burden without increasing head count and reducing the quality and purity of final registered research chemistries.

We at Virscidian have worked with multiple experts across the industry over many years to fine tune and produce an end-to-end purification workflow that allows you to purify compounds either as singletons or large numbers of plate-based samples.

High-Throughput Purification Process

HTP generally involves analyzing the samples by chromatography/mass spectrometry multiple times before the samples are ready to be stored in compounds libraries or used for downstream processing.

The first step of the process, often referred to as the PreQC or crude screening phase, is determining the conditions necessary to successfully separate the compound(s) of interest from other contaminants, maximize recovery, and minimize the volume and number of collected fractions.   Multiple parameters can be varied including mobile phase composition, pH, temperature, stationary phase, and even evaluating liquid chromatography vs. supercritical fluid chromatography.

After determining which chromatography method provides the best separation for each compound, a larger volume of each sample is injected onto preparative scale chromatography systems where the compound of interest is detected and automatically collected into vials or tubes using a fraction collector.

Depending on the workflow, the collected fractions might be analyzed again via chromatography/mass spectrometry to ensure the correct compounds were collected and determine their purity and in some cases their concentrations. This step is often called fraction QC.

If samples are dried down and re-solubilized after fraction collection, they are often re-analyzed via chromatography/mass spectrometry, making sure that the final samples are pure. This analysis, referred to as Final QC, is used for the registration of the purified material with compound management teams.

Drawbacks of Manual Data Processing

Although running the samples through the chromatography/mass spectrometry instrument can be automated and performed without manual intervention, historically it’s required significant human involvement to review these results.

Scientists must determine which chromatography method yields the best results for each compound and then determine how best to set up fraction collection triggers so that they collect as much of the compound of interest as they can in the lowest number of fractions.

This process can result in  multiple hours of data evaluation, and and lead to variable results based on the level of experience,  intuition, and preferences of each scientist.

Automation Increases Throughput

The benefits of automating the data review process are multifold and include higher throughput with fewer person-hours required for data review, and more reproducible results. In the past decade, several companies have begun developing software to automate this decision making, with Virscidian’s Analytical Studio leading the field.  

The impact of software-based automation can be extremely significant.  In one customer example, cycle times for an end-to-end purification workflow were reduced by more than 50% using Analytical Studio. In another example, using a review-by-exception approach facilitated by Analytical Studio, final QC screening efficiency improved by more than 400%.

Conclusion

Despite the challenges of data review, there are sizable benefits from performing HTP and therefore, it has become a valuable tool in the drug discovery process. It has made it possible to screen large libraries of compounds rapidly and efficiently for potential drug candidates, significantly accelerating the drug discovery process.  As the technology continues to develop and automated data processing tools become more widespread, it will become even faster, more efficient, and more affordable.