#Japan #Tsukuba #In-Lab LLM #Humanome Lab #Nabe International

In-Lab LLM Starter Kit: A New Era for Life Sciences Research

The collaboration between Humanome Lab and Nabe International is set to transform life sciences research through the introduction of the In-Lab LLM Starter Kit. This innovative AI analysis platform caters specifically to bio-pharmaceutical and clinical research, allowing researchers to efficiently process high-security data within a secure laboratory environment. Starting September 1, 2025, the In-Lab LLM Starter Kit will be available for a special price, with only five units offered.

Revolutionizing Research with AI

The usage of cloud-based AI tools like OpenAI's ChatGPT and Google's Gemini in research has been heavily restricted due to security concerns surrounding sensitive data. However, the In-Lab LLM Starter Kit allows for significant advancements while keeping all interactions contained within a closed server environment. The platform operates with a conversational model that boasts capabilities akin to ChatGPT, but it is entirely offline. Researchers can analyze genes, retrieve clinical information, and examine new hypotheses without ever compromising their data.

The In-Lab LLM Starter Kit provides support throughout the research process, from generating hypotheses and automatically creating analysis codes to interpreting results comprehensively. By providing this unique solution, the kit promises to not only accelerate research cycles but also minimize operational risks associated with data management.

The Background Behind Development

In recent years, generative AI has become an indispensable tool in scientific research, assisting with hypothesis generation, statistical coding, and result interpretation. However, concerns about the potential misuse of sensitive information in cloud-based LLM services have inhibited their broad adoption, particularly in fields that handle confidential genomic and clinical data. The inception of OpenAI's open-weight model, gpt-oss, has offered a viable solution for operating high-performance language models securely within research facilities.

By harnessing advanced AI development from Humanome Lab and Nabe International's robust IT infrastructure expertise, the In-Lab LLM Starter Kit was born. This initiative aims to facilitate breakthroughs in life sciences research by overcoming the barriers posed by traditional cloud-based AI systems.

Addressing Security Concerns with Local Solutions

Research environments have seen a surge in the adoption of generative AI for tasks such as summarizing research papers and generating statistical analysis codes. However, the challenge has always been the restriction on entering sensitive raw data into external cloud systems. With the In-Lab LLM Starter Kit, researchers can input their data directly from a secure browser interface and engage in hypothesis generation and analysis in a conversational manner. This seamless integration greatly enhances the research development cycle.

Key features and advantages of the In-Lab LLM Starter Kit include:

• - Hypothesis Generation: Inputting research data prompts the model to suggest new hypotheses and guide further inquiries through an interactive dialogue.
• - Automatic Code Generation: The platform formulates the required analytical pipelines—such as scripts and commands—necessary to validate the proposed hypotheses interactively.
• - Instant Execution and Outcome Interpretation: Researchers can execute the generated codes in real-time, discussing any issues or interpretations directly with the In-Lab LLM. This rapid feedback loop considerably reduces the trial-and-error phase of research.

Unique Advantages of the In-Lab LLM Starter Kit

Planning for future expansions, the Starter Kit is designed to support research labs in creating their own LLM and analysis environments. An announcement about collaborations with bioinformatics tools and support for larger scale analyses is expected on September 3, 2025, during the 13th Combined Conference on Life and Pharmaceutical Informatics at Nagoya University.

Security and Accessibility

The In-Lab LLM Starter Kit is designed to handle sensitive genomic and clinical data securely, without relying on external cloud services. The system operates entirely within a dedicated local server, ensuring that all critical information remains confidential.

Configured to run gpt-oss-20b comfortably, the kit is accessible at a price that fits within the budgets of research labs. Users can comfortably conduct hypothesis development, automatic code generation, and research idea organization within a safe, local environment.

Easy Browser Access

Equipped with a standardized Docker container environment, the In-Lab LLM is easily accessible via web browsers, making it convenient for all members of a research team. No need for additional accounts for every user means labs can scale their use flexibly according to their needs.

Flexible Options for Researchers

The kit also provides options for supporting research-specific needs, including:

• - RAG Construction Support: Incorporate documents and analytical data from the research lab into the private LLM environment, enabling responses that leverage specialized knowledge.
• - Fine-Tuning Support: The Humanome Lab will assist in developing a custom LLM tailored for specific research objectives, alongside enhancing the generative AI model.
• - Hardware Specification Expansion: As research goals evolve, labs can enhance computational and storage capacities, as well as integrate additional analysis tools and advanced models like gpt-oss-120b.

Offering Various Plans

Three types of plans will be available based on the research phase and objectives, all limited to five units:
1. Starter Kit: A private LLM operating in a completely enclosed environment that significantly enhances the efficiency of daily research tasks, including paper summarization and programming assistance.
2. LLM Expansion Support Plan: A custom environment that incorporates lab documents and analysis data into the private LLM, supported from consultation to deployment.
3. Hardware Specification Expansion Plan: Options available for those needing higher performance LLM capabilities or additional analytical server functions.

Supporting Researchers with Confidence and Efficiency

Remarks from Company Leaders

Jun Se-se, CEO of Humanome Lab: “Our mission is to leverage AI to address new challenges in life sciences and contribute to healthier living. Data security is crucial in life sciences and drug discovery. With the In-Lab LLM Starter Kit, we aim to create an environment where researchers can concentrate on their work without data management concerns.”

Nobuyuki Watanabe, CEO of Nabe International: “For years, we have focused on building genomic data analysis environments for life sciences researchers. The In-Lab LLM Starter Kit is a significant undertaking that merges Humanome Lab's expertise in AI with our reliable IT infrastructure. We view this as a promising step to provide researchers with a secure environment tailored for life sciences research.”