2026 Drug Target Landscape Expands Across New Modalities

The drug target landscape has changed dramatically since 2000, as genomics, proteomics, computational science, automation, and new therapeutic modalities have expanded what researchers can pursue.

A new Nature Reviews Drug Discovery analysis maps 25 years of progress across 686 mechanism-of-action biomolecules and 1,702 approved drugs, revealing both the expansion of druggability and the continued concentration of success around established target families. 

Image

A 25-Year Map of the Drug Target Landscape

The new review by Halip, Avram, Overington and colleagues examines the evolution of therapeutic targets from 2000 to 2024, with additional analysis of new-drug approvals across the U.S. Food and Drug Administration (FDA), European Medicines Agency (EMA), and Japan’s Pharmaceuticals and Medical Devices Agency (PMDA) through June 2025.

Published in Nature Reviews Drug Discovery on 21 September 2026, it builds on earlier landmark target-landscape studies from 2006 and 2017. 

The central dataset contains 686 biomolecules modulated by 1,702 approved drugs.

Those figures need some qualification.

Not every protein that has ever been investigated as a potential target. A mechanism-of-action (MoA) target is defined as a protein or other biomolecule directly bound by a drug to produce its therapeutic effect.

The analysis excludes targets associated only with ADMET processes or targets without established clinical relevance. 

That makes the dataset particularly useful for understanding clinically validated druggability, rather than theoretical target space.

The 686 biomolecules therefore represent a much narrower population than the thousands of proteins catalogued in resources such as IUPHAR/BPS, which list more than 3,000 human protein targets across a broader pharmacological database. 

This is important because a protein can be biologically interesting without being clinically validated; it can be genetically associated with disease without having a suitable binding site, and it can bind a molecule without producing a useful therapeutic effect. Or it can be technically druggable but impossible to reach safely in patients.

The Nature review focuses on the much smaller intersection where biology, pharmacology, therapeutic modality, and clinical evidence meet.

What the 1,702-drug dataset reveals

The review does more than count targets.

It examines how the target landscape has changed alongside:

• Drug modality
• Protein target class
• Therapeutic area
• Drug potency
• First-in-class status
• Target novelty
• Regulatory approval patterns
• The emergence of new drug discovery approaches

This creates a longitudinal view of the drug target landscape rather than a static target catalogue. 

It also connects directly to the evolution of databases such as DrugCentral, which integrates molecular structures, drug information, clinical effects, and pharmacological relationships. The authors cite DrugCentral as one of the key resources underpinning this type of target analysis. 

The resulting picture is not simply one of more targets, but, rather, it’s a story about more ways to reach biological targets.

This sort of framing is growing in importance as drug discovery moves beyond conventional small molecules.

Established Drug Target Families Still Dominate

The expansion of druggability has not erased the importance of established target families.

GPCRs, enzymes, ion channels, nuclear hormone receptors, catalytic receptors, transporters, and other proteins remain central to modern pharmacology.

The review analyses these protein groups across therapeutic areas and historical periods, showing how their contribution to approved medicines has evolved.

The paper’s target-class analysis should also be interpreted carefully when using the headline numbers.

The figures cited in preliminary discussions of the paper include:

• 423 GPCR-related classifications, 
• 361 enzyme classifications, 303 other-protein classifications, 
• 169 ion-channel classifications, 
• 164 nuclear-hormone-receptor classifications, 
• 158 catalytic-receptor classifications, 
• 114 transporter classifications, and 
• 65 endogenous-peptide classifications.

However, these figures should not be described as eight sets of unique targets. Their combined total is greater than 686, so they represent class-level associations within the paper’s analysis rather than eight mutually exclusive counts of the 686 biomolecules.

The more defensible research headline is therefore the authors’ stated 686 MoA biomolecules modulated by 1,702 drugs. 

The class-level data nevertheless illustrate an important point.

The drug target landscape, and drug discovery in general, remains heavily anchored in target families with decades of pharmacological knowledge.

That does not mean the industry has stopped discovering new biology. It does mean that translating biological insight into medicines still favours targets for which researchers can establish a convincing relationship between molecular intervention and therapeutic effect.

GPCRs are a useful example. They represent one of the most extensively exploited target families in pharmacology. Yet the modern GPCR field is also expanding beyond traditional orthosteric ligand binding.

Structural biology, allosteric modulation, biased signalling, computational modelling, and increasingly sophisticated ligand-design approaches are opening additional ways to manipulate these receptors. The review cites recent GPCR drug-discovery work as part of this broader evolution. 

The same principle applies to enzymes.

Enzyme targets remain fundamental because their catalytic mechanisms can provide defined intervention points. But the modalities used to influence them are changing, while previously difficult biological mechanisms are becoming more accessible through advances in chemistry and structural biology.

This is one reason the target landscape cannot be separated from modality innovation.

A target that was inaccessible to one therapeutic format may become clinically actionable when another format becomes available.

Image

How Are Biologics Reshaping What Counts as Druggable?

One of the clearest trends identified in the review is the growing contribution of biologics.

For this analysis, the authors define biologics broadly to include antibodies, oligonucleotides, other proteins, enzymes, fusion proteins, and peptides. Vaccines, gene therapies, and cell therapies are excluded from this particular definition.

That methodological choice is important because the growth of biologics changes the boundaries of what pharmaceutical companies can realistically target.

A conventional small molecule generally needs a suitable pocket or binding surface that can be reached with an appropriately sized chemical structure.

An antibody operates differently.

It can recognise a large extracellular surface with high specificity; an oligonucleotide can act through nucleic-acid sequence recognition; a peptide can reproduce or interfere with biological signalling, and a fusion protein can combine targeting and effector functions.

The modality changes the accessible biology.

The analysis shows how this has played out over the past quarter-century. According to the review’s senior co-author Sorin Avram, endogenous peptide targets have increased 7.2-fold since 2000, with the growth associated with antibody therapeutics, while biologics and small molecules reached approval parity in 2023.

That is a major structural change in pharmaceutical innovation and means that the modern discovery organisation cannot assess a target solely through the lens of medicinal chemistry.

A target assessment increasingly needs to ask:

• What modality can reach it?
• Where is the target expressed?
• What biological interaction should be altered?
• Can the intervention achieve sufficient selectivity?
• Can the resulting therapeutic index support clinical development?

These questions connect target identification directly to platform capabilities.

The expansion of modalities also helps explain why the industry’s definition of the “druggable” proteome continues to evolve. A protein that looks difficult for a conventional small molecule may become accessible through an antibody, degraders, an oligonucleotide, a targeted conjugate, or another modality.

The discovery challenge therefore moves from simply identifying what to target towards determining how to intervene successfully.

That shift is particularly important for research teams investing in new discovery platforms.

A target-identification engine can generate thousands of biological hypotheses. The commercial value comes from identifying those that can survive the next stages of validation, modality selection, chemistry or engineering, pharmacology, safety assessment, and ultimately human testing.

This is where the historical data in the Nature review becomes strategically useful.

The 686 clinically relevant MoA targets are not merely a list of successful proteins. They represent a record of what the pharmaceutical industry has managed to convert from biological hypothesis into therapeutic intervention.

The next frontier may therefore depend less on finding ever larger numbers of hypothetical targets and more on expanding the set of biological mechanisms that can be reliably validated, reached, modulated, and translated into medicines.

View interview How Should Pharma Companies Audit AI Vendors?

How Should Pharma Companies Audit AI Vendors?

“This is one of the riskiest industries there is because you're dealing with human lives, you're dealing with experimental protocols, and you're dealing with regulatory bodies where you might not get another shot at that clinical trial."

Image

Why Are First-In-Class Targets Still Hard to Find?

The expansion from established targets to new mechanisms is one of the most important aspects of the drug target landscape painted in the review.

A larger target landscape does not automatically mean that drug discovery has become easier. Moving from a biologically interesting protein or biomolecule to a clinically validated mechanism still requires evidence across pharmacology, disease biology, safety, drug exposure, and human outcomes.

The review defines a first-in-class drug as one whose mechanism-of-action target had not previously been modulated by an approved drug. This provides a useful way to separate genuine expansion of the target landscape from continued innovation around established biology.

A large proportion of pharmaceutical innovation can occur without introducing an entirely new target. Existing targets can support new binding sites, new selectivity profiles, new modalities, different tissue exposure, combination strategies, or improved pharmacology.

Kinases provide a particularly clear example. The authors report 136 approved kinase inhibitors across 67 kinase targets. This is a substantial concentration of successful drug discovery around one protein superfamily, particularly in oncology, where kinase signalling has become deeply embedded in therapeutic development.

The kinase story therefore illustrates two different forms of innovation. One involves discovering and clinically establishing new targets, while the other involves progressively improving how an established target can be modulated.

For drug discovery teams, both matter.

A portfolio focused only on entirely new targets can carry substantial biological and translational uncertainty. However, a portfolio focused only on established targets can face intense competition and diminishing differentiation.

The strategic challenge is to understand where target novelty creates sufficient therapeutic opportunity to justify the additional development risk.

Target Expansion Is a Modality Story

The review also shows that the definition of a druggable target is increasingly connected to the modality used to reach it.

Small molecules historically shaped the target landscape because they could be designed to interact with defined pockets and were often compatible with oral delivery. Biologics, peptides, oligonucleotides, and other modalities operate under different physical and biological rules.

The growth of endogenous peptide targets demonstrates this change particularly clearly. According to the authors, endogenous peptide targets increased 7.2-fold since 2000, with antibody therapeutics identified as a major driver of the expansion.

That is more than an astonishing statistic. It signals a shift in what pharmaceutical research can practically engage.

Antibodies can recognise larger or more complex molecular surfaces than many conventional small molecules, while oligonucleotides can influence RNA biology.

Protein-based approaches can exploit extracellular or highly specific biological interactions, and other emerging modalities can change the relationship between target selection and conventional notions of binding-site druggability.

The review also reports that biologics and small molecules reached approval parity in 2023. That point is important when interpreting the industry’s changing target landscape. Therapeutic innovation is no longer adequately described through the lens of small-molecule chemistry alone.

Instead, target selection increasingly needs to be considered alongside modality selection.

A target may be unattractive for one modality but accessible through another. Conversely, a biologically compelling target may still be difficult to translate if the selected modality cannot achieve the required tissue distribution, exposure, selectivity, durability, or safety profile.

This creates a more integrated drug discovery problem. Target identification, modality selection, molecular design, delivery, and translational assessment increasingly need to operate as connected decisions rather than sequential activities.

View interview AI in Pharma: Why the Future of Healthcare Starts With Patients, Not Tech

AI in Pharma: Why the Future of Healthcare Starts With Patients, Not Tech

Kate O’Reilly, President & Chair of the Healthcare Businesswomen’s Association (HBA) Dublin-Ireland Chapter & Healthcare Transformation Partner at Roche, discusses AI in pharma, patient engagement, and the future of healthcare innovation.

Image

What Is the Geography of Target Innovation?

The review also provides a geographic dimension to the evolution of drug targets. The senior author Sorin Avram reports that the U.S. accounted for 78% of novel target validation since 2000.

This statistic does not mean that target discovery occurs exclusively in the U.S. Rather, it highlights the concentration of clinically successful target innovation within the U.S. regulatory and pharmaceutical ecosystem over the period analysed.

The distinction between discovery and regulatory confirmation is important. A target can originate in academic research, biotechnology, international pharmaceutical programmes, or publicly funded science, while the eventual therapeutic programme may be developed and approved elsewhere.

The geographic pattern therefore raises broader questions about where target hypotheses originate, where they are developed, which regulatory systems confirm them, and how intellectual property, capital, clinical infrastructure, and translational expertise influence the movement from biological discovery to approved medicine.

For drug discovery teams, this creates a strategic intelligence issue. Tracking emerging targets requires monitoring approved medicines and visibility across academic research, biotechnology pipelines, clinical trials, licensing activity, modality platforms, and regulatory decisions.

The 686-target landscape is therefore better understood as a dynamic network than as a static catalogue.

AI Will Change the Next Drug Target Expansion Cycle

The timing of this review is significant. It arrives alongside a second 2026 Nature Reviews Drug Discovery review focused specifically on target identification and assessment in the era of AI.

That review describes AI as increasingly important for analysing large datasets and complex biological networks and examines applications spanning target exploration, cellular imaging, model-based assessment, and experimental workflows.

It also emphasises that AI approaches have limitations and require appropriate assessment rather than treating computational prediction as equivalent to biological or clinical proof.

The two reviews therefore fit together.

The evolving target landscape describes what has happened to therapeutic targets over the past quarter century, while the AI review examines how increasingly sophisticated computational approaches may affect the process used to identify and assess the next generation of targets.

This distinction is critical. AI can expand the search space, integrate heterogeneous evidence, identify relationships across biological networks, and help prioritise hypotheses. It does not remove the need to establish whether a target has a meaningful role in human disease or whether modulation of that target can produce an acceptable therapeutic window.

That creates an important opportunity for R&D teams. If AI can improve the ranking of target hypotheses before significant experimental investment, organisations may be able to investigate a broader biological space without treating every hypothesis as an equivalent development opportunity.

The value of AI may therefore lie less in simply generating more targets and more in improving the quality, integration, and speed of decisions made around them.

This connects directly with the growing role of computational approaches in modern drug discovery.

Pharmatica’s Insights into the gap between AI adoption and measurable drug discovery impact examines why computational capability does not automatically translate into clinical outcomes.

The drug target landscape review adds another layer to that discussion in that the ultimate measure of progress is not how many hypotheses a system can generate, but how effectively the industry converts biological hypotheses into differentiated medicines.

Image

What Does the Expanding Target Landscape Means for Pharma R&D?

The 686 biomolecules and 1,702 approved drugs mapped by the review provide a useful baseline for understanding modern drug discovery.

The headline number is important, but the distribution behind it is arguably more informative.

Established target families continue to account for a substantial proportion of successful pharmacology. Kinases demonstrate how deeply a target class can be exploited, with 136 approved inhibitors spanning 67 targets.

At the same time, the 7.2-fold increase in endogenous peptide targets demonstrates how new modalities can expand the practical boundaries of druggability.

The 2023 parity between biologic and small-molecule approvals reinforces the same point from a different direction. Modality diversification is no longer peripheral to the target landscape. It is part of its evolution.

The next phase of drug discovery will therefore depend on connecting several layers of evidence: Human genetics, disease biology, molecular mechanisms, target tractability, modality suitability, pharmacology, safety, clinical translation, and regulatory precedent.

The review’s 2000–2024 analysis, supplemented by approval data from the FDA, EMA, and PMDA through June 2025, provides a foundation for monitoring that transition. It also shows why target intelligence cannot be reduced to a list of proteins associated with disease.

A meaningful target landscape needs to capture which targets have been clinically demonstrated, how they have been modulated, which modalities have succeeded, where novelty is emerging, and where established biology is still producing new therapeutic opportunities.

The Next Drug Discovery Frontier

The next 25 years of drug discovery are unlikely to be defined by a single technology or target class. The drug target landscape mapped by Halip, Avram, Overington, and colleagues points instead towards a progressively more diverse therapeutic ecosystem.

Human biology is generating more potential hypotheses; data science is making those hypotheses easier to connect; AI is changing how evidence can be integrated, and new modalities are changing which biological structures can be therapeutically addressed.

Yet the central challenge remains unchanged: Turning biological understanding into medicines that work in patients.

The expanding target landscape provides a quantitative record of how far that process has already moved.

At Pharmatica, we believe that the next opportunity lies in using that historical evidence to make better decisions about where Therapeutic Drug Discovery should move next.

Pharmatica: Insight. Connection. Impact.