A key resource for understanding gene regulation
JASPAR is an open-access, expertly curated database of transcription factor (TF) binding profiles. These profiles, which are position frequency matrices (PFMs) derived from experimental data, describe how TFs recognise specific DNA sequences across vertebrates, plants, fungi, insects, and nematodes. Because JASPAR is non-redundant, open-access, and highly trusted it has become a critical resource for regulatory genomics and modelling of gene regulation.
Expanding the classical PFM motif collections
The 2026 update of this long-lasting resource, established in 2004, expands the CORE collection with 306 new profiles (a 12% increase) and adds 433 new entries to the UNVALIDATED collection (a 60% increase). In addition, this new release features transcription factor binding site predictions and genome tracks for eight species and new computational tools for regulatory genomics, strengthening its usefulness across diverse research fields.
Introducing the JASPAR Deep Learning Collection
Traditional motif models capture essential TF–DNA interactions, but they simplify the biological reality. Deep learning approaches can recognise more complex patterns, such as the combination of multiple motifs.
With this update, the team fulfills a long-anticipated goal: the launch of the JASPAR Deep Learning Collection. This new resource includes 353 motif patterns for 240 TFs, summarising the binding features captured by deep learning models developed by collaborators in Anshul Kundaje’s group in Stanford.
These advanced models, available to the research community, pave the way to an improved understanding of how TFs read the genome with potential to uncover how DNA sequence variation shapes regulatory mechanisms, increasing the understanding of their influence on human health and disease.
A collaborative legacy
The first JASPAR release in 2004, created by Albin Sandelin, Wyeth Wasserman, and Boris Lenhard, laid the groundwork for what is now a major international initiative. For the 2026 release, research groups across ten countries have contributed to its development, reflecting two decades of collaborative effort.
Original publication
Baydar et al. JASPAR 2026: expansion of transcription factor binding profiles and integration of deep learning models. Nucleic Acid Research 2025. https://doi.org/10.1093/nar/gkaf1209
Link to JASPAR 2026