N-iX will build and support the AI, data engineering and LLM infrastructure behind COGNANO's antibody drug discovery research, across more than 300 million antibody gene reads and 30 million labeled sequences.
N-iX, a global technology partner for Pragmatic AI Software Engineering, has signed a Memorandum of Understanding (MoU) with COGNANO, Inc., a Kyoto-based biotechnology company. The agreement was signed at N-iX's office in Lviv.
COGNANO's VHH antibody data and IBMET technology
COGNANO has built a large-scale VHH antibody dataset. VHH antibodies are small antibodies that can detect subtle structural differences in disease-related molecules. The dataset includes more than 300 million antibody gene reads and 30 million labeled sequences across 50 antigen types.
Using its own technology, IBMET (Inverse Biomarker Exploring Technology), COGNANO combines statistics and large language models (LLMs) to analyze antibody data, disease-specific biomarkers and identify new drug targets. The dataset and method have been described in peer-reviewed research (see Supporting research below).
N-iX's role in AI and data engineering
As COGNANO's research grows, so does the volume of data its scientists need to process. Under the MoU, N-iX will provide support in AI, data engineering and LLM infrastructure for processing complex biological data at scale and extracting insights for COGNANO's research.
The initiative is being developed with support from the Japanese Services Group (JSG) at Deloitte Poland.
Photo (left to right): Wataru Takahashi, Partner Associate, Deloitte Poland; Sergii Lesniak, VP Business Development East Asia, N-iX; Akihiro Imura, President and CEO, COGNANO; and Taras Petriv, Customer Engagement Director, N-iX. Lviv, Ukraine, September 2026.
COGNANO perspective
Antibody sequence data is becoming a powerful language for decoding molecular recognition. For SEPIQ, COGNANO provided 1.8 million HER2 VHH sequences. With N-iX, we will integrate VHH big data, structural biology and LLMs to bidirectionally translate antibody-epitope correspondence and discover disease-specific molecular surfaces for selective therapeutics and diagnostics at scale.
N-iX perspective
COGNANO has built a highly specialized antibody dataset and a scientific platform around it. Our role is to make sure the AI and data infrastructure can support the scale and complexity of that research. This cooperation brings together COGNANO's expertise in antibody discovery with N-iX's experience in AI and large-scale data engineering.
Supporting research
COGNANO's antibody datasets and research relevant to this cooperation have been documented through company-published data and scientific publications:
- COGNANO reports 15 years of VHH antibody research, data covering 50 antigen types, more than 300 million antibody gene reads, and 30 million labeled antibody sequences. COGNANO research and technology overview
- AVIDa-hIL6: A Large-Scale VHH Dataset Produced from an Immunized Alpaca for Predicting Antigen-Antibody Interactions. Describes COGNANO-developed dataset containing 573,891 antigen-VHH pairs with binding or non-binding labels and 30 IL-6 mutants in addition to wild-type IL-6. NeurIPS 2023 paper – AVIDa-hIL6
- Inverse Biomarker Exploring Technology (IBMET): Mathematical Identification of Cancer-Specific Epitopes and Novel Targets (Biomarkers) from Big Data of Single-Domain Antibodies Recognizing Higher Structures. Cancer Research, 2024. AACR conference abstract
- Efficient discovery of an agonistic anti-OX40 nanobody by epitope-directed approach to address enrichment-driven epitope bias. Includes COGNANO-affiliated researchers and describes the use of next-generation sequencing and computational analysis in nanobody discovery. Journal for ImmunoTherapy of Cancer, August 2026 - anti-OX40 nanobody research
About COGNANO, Inc.
Headquartered in Kyoto, Japan, COGNANO uses antibody data and AI technology to develop highly accurate drug-discovery seeds and diagnostic tools. Drawing on its extensive datasets, the company efficiently identifies promising therapeutic candidates and significantly speeds up drug discovery. Learn more at cognanous.com.
