We've updated our Privacy Policy to make it clearer how we use your personal data. We use cookies to provide you with a better experience. You can read our Cookie Policy here.


High-Throughput Screening: Advances, Applications and Combined Approaches

A row of pipettes filled with red liquid
Credit: iStock
Listen with
Register for free to listen to this article
Thank you. Listen to this article using the player above.

Want to listen to this article for FREE?

Complete the form below to unlock access to ALL audio articles.

Read time: 12 minutes

High-throughput screening (HTS) is a drug discovery approach that involves screening vast libraries of small molecules to identify “hits” with therapeutic potential. As a result of exciting past scientific breakthroughs, HTS approaches have been miniaturized, standardized and automated. This, alongside more recent advances in cell culture and data analysis, has led to faster more reproducible screening with targets that more accurately reflect in vivo physiology.

In this article, we highlight various HTS strategies being used to interrogate large libraries of compounds and hear from researchers working to further streamline specific approaches to increase speed and improve quality and accuracy.

Let’s start by considering different approaches to HTS.

In vitro biochemical assays

A purified target (e.g., enzyme, receptor or hormone) is screened against a compound library. The readout is typically achieved using optical detection methods such as absorbance, fluorescence or luminescence.

Considerations: These assays can identify hits and quantify binding affinity but cannot reproduce a phenotypic or functional response. They are unable to provide information on cellular context.

Cell-based assays

Increasingly cell-based assays are used to provide a phenotypic readout, e.g., a change in signaling measured by a reporter gene or an effect on cell growth.

Considerations: They provide information on how compounds affect function in a cellular environment. However, drug compounds can sometimes fail to perform in this setting if 2D monolayer cultures are used, as they lack the native tissue microenvironment1 that influences cell behavior and function. These assays do not provide information on how the compound is exerting its effect – i.e., what the cellular target is, or whether it is hitting the desired target or not.


RNA interference (RNAi) screening

RNAi screening can be used for high-throughput analysis of gene function on a genome-wide scale.2

Considerations: This approach provides insights into functional networks in cells that could be targeted in drug discovery. It can be combined with high-content imaging.

These approaches are key in the early stages of drug discovery, offering researchers a diverse set of tools that can be used in combination depending on the specific requirements of the project and starting materials (i.e., libraries, X-ray structures of targets, available disease models). They each have their own merits and limitations, but all have universally benefitted from several key technological advances (Table 1).

Table 1: Key advances in assay technologies for HTS


HTS was made commercially feasible by the development of automated liquid handling workflows that make it time and cost-efficient to process thousands of parallel samples. Today’s automated workflows require limited supervision or manual intervention and improve screening quality and reproducibility.3,4

Microfluidic systems

Miniaturized microfluidics devices make it possible to screen compounds in high-density arrays, using small volumes of purified target and compound.5 This has increased the rate of screening and reduced the amount of reagents required. Microfluidic channels also provide a basis for building different connected compartments to mimic the interconnected nature of the human body.  


3-dimensional (3D) cell culture

3D cell culture has enabled tissue spheroids or miniature organs (organoids) to be derived from cell lines, primary cells, or stem cells which further enhance drug screening by better recapitulating the human physiological environment compared to 2D cell lines or animal models.6


High content screening

The addition of automated and/or live cell imaging methods alongside quantitative assays now makes it possible to capture more detailed phenotypic outputs in high throughput, by taking multiple views of a well under different light wavelengths and by labeling different targets within the individual cells.7 This can provide insights into cellular expression changes, or morphology changes at the organelle or single-cell level. Multiple properties of individual cells can also be studied simultaneously.

HCS assays are typically conducted using 2D monolayer cultures to achieve higher throughput. However, 3D cell culture systems (e.g., spheroids, organoids) can also be used.

Gene editing

Precise gene knockout using tools (e.g., CRISPR-Cas9) offer the potential to screen cells for targets that can then be pursued therapeutically.2,8 As CRISPR guides are very easy to synthesize, you can design customized libraries to knock out every gene in the genome or a particular set of genes, making the approach amenable to HTS.

DNA-encoded libraries

Advances in genetic sequencing allow compounds to be DNA barcoded, so that large compound libraries can be screened within a single microplate well, accelerating hit identification and optimization.9,10


Despite these advances, researchers are still working to optimize conventional HTS approaches and develop platforms and software to combine different methods. In the rest of this article, we’ll explore some examples of innovation in screening methods.

Navigating challenges in HTS


The practical logistics of screening compound libraries that can contain as many as three to four million compounds remains a major challenge in HTS. Physically screening each of these compounds against a single target in replicate is a time-consuming and labor-intensive undertaking, even with automation. Using more focused libraries for well-characterized targets is one option, but another route is to use mixtures of compounds against a single target. To achieve this, compounds can be attached to solid substrates such as beads to allow the purification of “hits”, but this is still not practical for the largest libraries. This has led to the introduction of DNA barcoding to generate DNA-encoded libraries.

DNA-encoded libraries

A short piece of DNA called a “headpiece” is split into portions and modified with different small molecules.9 A set of unique DNA barcodes is added to each portion of DNA, with each barcode corresponding to a specific small molecule. The resulting DNA-chemical conjugates are then pooled, split, chemically diversified and encoded with another set of DNA barcodes. This split-modify-encode-pool is usually repeated 2‒4 times and allows the generation of many thousands of DNA-encoded small molecules that can be screened against a protein target. 

Dr. Basilius Sauter is a postdoctoral researcher in Prof. Dennis Gillingham’s lab at the University of Basel. Sauter specializes in the areas of DNA-encoded libraries and DNA-modifying drugs.

“Although the concept of DNA barcoding is not new, it was only with advances in next-generation sequencing that DNA-encoded libraries became viable for larger library sizes,” he said. “Since then, the number of different chemistries available has rapidly increased, allowing the generation of more diverse DNA-encoded libraries covering a greater chemical space.”

Optimizing DNA-encoded libraries for HTS

One limitation of DNA-encoded libraries is that, unlike conventional HTS, the assay can only detect binders or non-binders, and you need to be able to separate these.  The conventional method is to use an affinity purification step, where you attach your target protein to a solid phase and wash the non-binders away. But this has the disadvantage that you can lose compounds that bind weakly which, with further optimization, could have been potential drug candidates.


To circumvent this, Sauter and colleagues have recently developed a new technique using an enzyme called terminal deoxynucleotidyl transferase (TDT).

“TDT is a very special polymerase because it can synthesize DNA without a template,” explained Sauter. “We made a fusion of TDT with our target protein of interest, such that when a DNA-encoded library member binds to the protein of interest, TDT is in close proximity. Then, when we incubate with only dATP, the TDT preferentially extends the 3’ terminus of the DNA barcode on the hit molecule and makes a long poly dA tail.”

This makes it possible to select hits in solution, which is an advantage when working with drug targets that start to denature or behave differently when on a solid phase for the affinity purification step. The polyA tail is also proportionate to the binding affinity of the compound, because it relies on the proximity of the drug to the target.11

Perhaps the most exciting and surprising result, however, was that the TDT method not only identified all the binders that were determined using the standard affinity purification step, but also found additional weak binders that had evaded previous screens.

“When we tested it, it turned out to be a very weak inhibitor of our target molecule,” said Sauter.” One of the issues with DNA-encoded libraries is they often identify hits that are strong binders but it’s not possible to take them forward because they aren’t amenable to lead optimization by medicinal chemists. We want to build on the work we’ve done so far to find a screening assay that is sensitive for these weak binders, especially for difficult targets where having a weak interaction is better than none.”

Combining different screening approaches and considering alternatives

Selecting a specific HTS approach often involves a compromise between using a method that provides fundamental information about drug binding characteristics to determine (e.g., specificity and sensitivity) and gaining functional insights about whether a drug’s action is achieving the desired phenotypic results. To solve this problem, a team led by Pavel Levkin at the Karlsruhe Institute of Technology (KIT) in Germany has developed a combined high-content, high-throughput platform that enables unified on-chip chemical synthesis, characterization and biological screening.12


The chip was developed as part of a European Research Council Proof-of-Concept project to develop more affordable and faster cell screening experiments in diagnostics and personalized medicine. Levkin and colleagues developed dendrimer-based surface patterning that enables the generation of high-density nanodroplet arrays. Each of the 50,000 droplets on a plate function as an individual nanovessel. With these vessels it is possible to conduct combinatorial chemical synthesis and then run a range of analytical assays – from on-chip detection using methods such as infrared spectroscopy or matrix-assisted laser desorption/ionization mass spectrometry, to high-content cell-based screening. By bringing together the traditionally separate steps of high-throughput chemical synthesis, reaction monitoring, compound characterization and biological screening, the technique offers the potential to unify the different approaches used in early-stage drug discovery.

Another significant challenge in HTS is managing large compound libraries and this, together with more challenging targets, has led drug discovery researchers to adopt fragment-based drug discovery (FBDD) as an alternative to HTS. In FBDD you initially screen chemical “fragments” to identify those with promising functionality, and then build on the fragment “hit” either by adding other chemistries or combining several fragments to create a more potent lead compound. Pioneered in the late 1990s by Professor Steve Fesik as “SAR by NMR”, it now constitutes an essential toolkit for the biopharma industry.13 However, despite this sensitivity, NMR screening has other limitations – researchers screening libraries of thousands of compounds quickly find themselves with a data analysis bottleneck, having to visually inspect thousands of spectra.

To address this problem, Geerten Vuister, professor of structural biology at Leicester University and chair of the Collaborative Computational Project for NMR and his team have developed a computational pipeline of analysis tools to support NMR-based screening.14 It started with tackling a fundamental challenge – the complexity of a fragment library.

“Even relatively small fragment libraries contain several hundreds of compounds and each has its own NMR spectrum, so when you combine these molecules in a single mixture, you might get overlapping spectral regions and it becomes difficult to distinguish between them,” explained Dr. Luca Mureddu, who developed the pipeline. “Moreover, factors such as experimental error, degradation of library compounds and varying buffer requirements for different protein targets all contribute to considerable variability between spectra.”

This led them to develop tools and workflows that adjust for experimental factors and post-processing that deals with the variability of the data improving the hit rate identification. After running the algorithms of the workflow, which can take as little as seconds for a small library, to around ten minutes for a library of a few thousand compounds, users receive a set of scores indicating a level of confidence about whether a compound is a hit or not.

“The most important premise is that we don’t want to replace the researcher,” said Mureddu. “So, we don’t provide a binary yes/no answer to whether something is a hit.” Instead, based on a combination of scores, users can not only identify reliable positive or negative hits, but also easily inspect the data for compounds that fall into the “uncertain” zone, potentially prioritizing these for follow-up screening using another established method.

“Drug companies tend to use a variety of screening approaches, so it was important to develop a pipeline that users can customize to their own specific screening needs,” highlighted Dr. Vicky Higman, a Research Fellow in Vuister’s team specializing in NMR.

“We’ve already seen that our collaborators are increasing the size of their screening libraries, because they can now analyze data much more quickly and reliably.”

About the interviewees:

Dr. Basilius Sauter is a postdoctoral researcher in Professor Dennis Gillingham’s lab at the University of Basel. Sauter specializes in the areas of DNA-encoded libraries and DNA-modifying drugs.

Dr. Luca Mureddu is a postdoctoral researcher at the University of Leicester and part of the 
Collaborative Computational Project for NMR working group. Mureddu focuses on focus on developing software tools for fragment based-screening to enable and accelerate high quality and reproducible research. 


1.      Wei F, Wang S, Gou X. A review for cell-based screening methods in drug discovery. Biophys Rep. 2021;7(6):504-516. doi: 10.52601/bpr.2021.210042

2.      Taylor J, Woodcock S. A perspective on the future of high-throughput RNAi screening: Will CRISPR cut out the competition or can RNAi help guide the way? J Biomol Screen. 2015;20(8):1040-1051. doi:10.1177/1087057115590069

3.      Hansel CS, Plant DL, Holdgate GA, Collier MJ, Plant H. Advancing automation in high-throughput screening: Modular unguarded systems enable adaptable drug discovery. Drug Discov Today. 2022;27(8):2051-2056. doi: 10.1016/j.drudis.2022.03.010

4.      Green CP, Spencer PA, Sarda S. Advancing automation in compound management: a novel industrial process underpinning drug discovery. Drug Discov Today 2021;26(1):5-9. doi: 10.1016/j.drudis.2020.09.032

5.      De Stefano P, Bianchi E, Dubini G. The impact of microfluidics in high-throughput drug-screening applications. Biomicrofluidics. 2022;16(3):031501. doi: 10.1063/5.0087294

6.      Wang Y, Jeon H. 3D cell cultures toward quantitative high-throughput drug screening. Trends Pharmacol Sci. 2022;43(7):569-581. doi: 10.1016/j.tips.2022.03.014

7.      Blay V, Tolani B, Ho SP, Arkin MR. High-throughput screening: today's biochemical and cell-based approaches. Drug Discov Today. 2020;25(10):1807-1821. doi: 10.1016/j.drudis.2020.07.024

8.      Shalem O, Sanjana NE, Zhang F. High-throughput functional genomics using CRISPR-Cas9. Nat Rev Genet. 2015;16(5):299-311. doi: 10.1038/nrg3899

9.      Satz AL. What do you get from dna-encoded Libraries?. ACS Med Chem Lett. 2018;9(5):408-410. doi: 10.1021/acsmedchemlett.8b00128

10.  Reddavide FV, Cui M, Lin W, et al. Second generation DNA-encoded dynamic combinatorial chemical libraries. Chem Commun (Camb). 2019;55(26):3753-3756. doi: 10.1039/c9cc01429b

11.  Schneider LA, Sauter B, Dagher K, Gillingham D. Recording binding information directly into DNA-encoded libraries using terminal deoxynucleotidyl transferase. J Am Chem Soc. 2023;145(38):20874-20882. doi: 10.1021/jacs.3c05961

12.  Benz M, Asperger A, Hamester M, et al. A combined high-throughput and high-content platform for unified on-chip synthesis, characterization and biological screening. Nat Commun. 2020;11:5391. doi: 10.1038/s41467-020-19040-0

13.  Shuker SB, Hajduk PJ, Meadows RP, Fesik SW. Discovering high-affinity ligands for proteins: SAR by NMR. Science. 1996;274(5292):1531-1534. doi: 10.1126/science.274.5292.1531

14.  Mureddu LG, Ragan TJ, Brooksbank EJ, Vuister GW. CcpNmr AnalysisScreen, a new software programme with dedicated automated analysis tools for fragment-based drug discovery by NMR. J Biomol NMR. 2020;74(10-11):565-577. doi: 10.1007/s10858-020-00321-1