The Korea Drug Development Fund (KDDF) project list brings together drug modalities, research targets, development stages and participating institutions. Looking across the projects reveals how much of the portfolio is devoted to the early work of finding candidates.
We examined the 553 projects in KDDF Pipeline v5.0, downloaded from the official KDDF brochure page, from three perspectives. The brochure was published in February 2026 and describes projects as of December 2025. Our analysis covers only the projects included in that publication. [1]
372 of 553 projects are in discovery
Discovery is the early research that finds and refines potential treatments. Preclinical research evaluates a candidate’s properties before studies in humans. Clinical trials assess safety, efficacy and other outcomes in people; this publication includes Phase 1 and Phase 2 projects.
By development stage, the list contains 372 discovery projects, 116 preclinical projects, 45 Phase 1 projects and 20 Phase 2 projects. Discovery represents 67.3% of the total. These figures count projects at each stage; they are not the proportions of candidates advancing from one stage to the next. [1]
Figure 1. All 553 projects, grouped by development stage. Every percentage uses 553 as its denominator. Source: KDDF Pipeline v5.0, data as of December 2025.
Discovery and preclinical research together account for 488 projects, placing most of the listed work before clinical trials. The research topics visible in this portfolio therefore lean toward finding new candidates and evaluating their properties.
Small molecules account for 41.0%
A drug’s modality describes the type of material or therapeutic approach it uses. Small-molecule drugs use relatively small chemical compounds; antibody drugs use proteins that recognize particular targets.
Under the source’s classification, there are 227 small-molecule projects (41.0%) and 101 antibody projects (18.3%). Gene and nucleic-acid approaches, which use materials related to genetic information, account for 66 projects. Recombinant-protein approaches, using proteins produced through genetic engineering, account for 54. [1]
Figure 2. The four largest modalities and the remaining five combined. The total is 553 projects, retaining the source’s classifications. Source: KDDF Pipeline v5.0; counts recalculated by Nadoo Bio.
The remaining five categories contain 105 projects: 34 in targeted protein degradation, which seeks to remove particular proteins; 30 in cell therapy, which uses cells; 20 in other biological approaches; 14 in other synthetic approaches; and seven in radiopharmaceutical therapy, which uses radioactive substances.
The classification rules matter when interpreting these counts. This publication includes ADCs, or antibody–drug conjugates—antibodies linked to drugs—in the antibody category. Its antibody count may therefore differ from that of another source that treats ADCs as a separate modality.
University projects concentrate in discovery; company projects span several stages
Grouped by participating institution, the projects comprise 417 company-led projects, 85 university projects, 40 projects at government-funded research institutes or national agencies, and 11 hospital projects. These institution types were assigned for this analysis using the institution names in the source. [1]
Figure 3. Counts represent projects, not institutions. The clinical column combines Phases 1 and 2. Source: institution and development-stage entries in KDDF Pipeline v5.0; institution classification and aggregation by Nadoo Bio.
All 85 university projects in this list are in discovery. Companies have 237 discovery projects, 115 preclinical projects and 65 clinical projects. Government-funded institutes and national agencies have 39 discovery projects and one preclinical project; hospitals have 11 discovery projects.
All 85 university projects are in discovery, while companies account for 180 of the 181 preclinical and clinical projects. This table describes the stages of different projects. It does not track the transfer of university research to companies.
From a project list to a research question
The list can help identify research topics where computation may be useful. For a small-molecule discovery project, for example, one question could be which candidates to include in the next experiment.
Figure 4. An example of using project classifications to define what to compare, which inputs are available and which results are needed. Explanatory diagram by Nadoo Bio.
The required output is more than a score for each candidate. A useful comparison also records which structures were evaluated and whether rankings remain similar when conditions change. Public project lists may not include the structures of the actual candidate compounds, so performing such a comparison requires materials supplied by the researcher.
Early discovery is a substantial part of company research too
Companies conduct 237 of the 372 discovery projects. Looking only at the 417 company-led projects, more than half are in discovery. In this dataset, finding and refining early candidates makes up a large share of company projects as well as university projects. [1]
A subsequent article examines candidate comparison in early discovery. When different tools rank the same compounds differently, what evidence can inform the choice of the next candidate? We will look at an actual calculation example, comparing the structures and scores of candidates whose rankings changed.
Data and counting rules
We counted 553 distinct project numbers. Development stages follow the source, with discovery subcategories combined.
Modalities follow the source’s top-level categories. Each project is counted once in the modality assigned by the source.
Percentages divide the category count by 553 and round to one decimal place. Institution types are analytical classifications based on the participating institution’s name.
The counts describe the distribution of research projects. They do not address project budgets or whether computational tools were used.
References
Korea Drug Development Fund (KDDF). KDDF Pipeline, Version 5.0 brochure. Published February 2026; project information as of December 2025. Downloaded from the official KDDF brochure page. Counts are based on the summary chart on printed page 4 and detailed entries on printed pages 6–41.