When Genomes Converse with AI! Exclusive Reveal of the Birth Path of SynGears Omni, China's First 'Omics+' Enzyme Source Intelligent Agent

"Imagine you're a natural products researcher. When you receive sequencing data from a rare fungus, what awaits you is two months of shuttling between a dozen different software programs, command lines, and overseas databases..."

"Now, you simply ask SynGears Omni one question: 'Is there an enzyme in this species that could synthesize anticancer compound X?' Then go grab a coffee."

On June 26, AI biomanufacturing company SynBioLab announced a major upgrade to its Xuanzhu Computing Core Platform SynGears—adding multi-omics analysis modules and embedding an AI agent architecture on top of its existing protein design capabilities, successfully building China's first "Omics+" Enzyme Source Intelligent Agent: SynGears Omni.

The "Omics+" Enzyme Source Intelligent Agent

Deep gene mining and analysis using omics represents the "first mile" of biomanufacturing. SynBioLab's SynGears Omni empowers the platform with new "omics source" capabilities combined with AI agent orchestration, enabling not only fully automated parsing of genomic, transcriptomic, and metabolomic data, but also direct connection of analysis results to downstream enzyme mining, protein design, and target product synthesis pathway screening.

Additionally, for highly sensitive data involving core strains and proprietary sequences—industrial biological assets—the platform ensures data localization and autonomous control from the underlying architecture, directly addressing the industry's core pain points in data security and intellectual property protection.

Currently, most AI biomanufacturing platforms domestically and internationally focus primarily on modifying known proteins or optimizing known strains. SynGears Omni, upgraded from SynBioLab's SynGears platform, is the industry's first commercial intelligent agent to integrate "omics source mining" as a new capability with downstream biomanufacturing.

In essence, the platform achieves end-to-end fully automated integration from raw sequencing data to biomanufacturing industrial transformation for the first time, moving biomanufacturing from "fragmented innovation" to "source innovation," from "starting halfway" to "reaching products directly from the source."

Full-Process Omics: Conversation Yields Results

As AI agents in protein design and small molecules continue to emerge, integrating full-process omics with AI agents remains a deep water zone few have ventured into: existing platforms mostly focus on only protein or small molecule levels, always lacking an integrated platform that can simultaneously handle protein design and multi-omics data with natively embedded agent architecture.

It was against this backdrop that SynBioLab's computing team began pondering a question: Could complex omics workflows and existing protein computing capabilities be unified within the same agent framework?

SynGears Omni is the product of this thinking. It is not a traditional omics analysis platform in the conventional sense, but rather an upgraded version built upon the Xuanzhu Computing Core protein platform SynGears. By introducing omics capabilities and agent architecture, the platform can simultaneously understand omics data and protein information, thereby integrating workflows that were previously scattered across multiple software and research stages into a single system.

SynBioLab's computing team was candid in interviews: "Omics workflows are extremely complex, with intricate and cumbersome software dependencies. We started building from June-July last year and didn't complete the genomics workflow prototype until year-end..."

Omics software is scattered across different ecosystems—from container images to package managers to source code compilation—installation and dependency management itself is a barrier; while different species, genome sizes, and sequencing types require flexible switching of analysis strategies, further widening the gap between wet labs and dry labs in the 0-to-1 phase.

Currently, the industry has long relied on outsourcing services, but traditional methods still incur high omics analysis costs—large genome assembly costs can reach hundreds of thousands of yuan. Additionally, existing public databases are scattered and incomplete, and domestic access to omics databases like the U.S. National Center for Biotechnology Information (NCBI) is restricted, bringing instability to research.

"Our users were very excited after hearing that the SynGears platform upgraded its omics module—they wanted to get data on species they're interested in to try it out. We've also opened it to individual users for continuous system beta testing," SynBioLab's computing team stated in the interview.

On the other hand, omics is not an "optional component" in biomanufacturing, but rather the innovation source for de novo discovery of nearly all enzymes, proteins, and functional genes. Especially for non-model organisms and natural products, multi-omics data including genomics, transcriptomics, and metabolomics must be relied upon for de novo analysis; additionally, systematic optimization analysis of engineered strains cannot be separated from omics analysis drive.

Therefore, the industry urgently needs an end-to-end omics intelligent agent that can perform joint reasoning across genomics, transcriptomics, metabolomics, protein structure, and multiple layers, quickly locking onto target genes or key enzymes of interest, aiding precise design of target product biosynthesis pathways, thereby dramatically shortening the R&D cycle from data to product.

SynBioLab's computing team also sincerely expressed in the interview: "SynGears Omni is not only designed for conversational research scenarios—we've also made our backend technical route, including all thresholds, publicly available on the webpage. Users can adjust parameters based on suggested thresholds to achieve technical goals faster. Scientific research should never be closed—we hope SynGears Omni can provide a more open, more efficient collaborative ecosystem for everyone."

Conversing with the SynGears Omni assistant using natural language, the most direct perception SynGears Omni brings us is "making biological computing as accessible as opening a browser."

In the conversation, SynBioLab's computing team also vividly depicted such a scenario for us: suppose you're a researcher, and before you lies a newly sequenced non-model organism genome.

In the past, you often needed to simultaneously master knowledge in genomics, bioinformatics, protein design, chemistry, be familiar with a dozen software and complex command-line workflows, and coordinate multiple research groups for collaborative verification. And training a technical person capable of independently completing this entire process often requires years of time and hundreds of thousands of yuan in funding investment—these hidden costs even exceed the direct expenses of individual projects. Even with personnel in place, the entire analysis process often takes a year or more to possibly complete.

On the SynGears Omni platform, you only need to ask one question in natural language: "Is there an enzyme in this species that could synthesize anticancer compound X?" Then go grab a coffee, and the platform can automatically complete genome parsing, candidate enzyme screening, and joint reasoning.

In a real-world test case of fully automated assembly and annotation of a eukaryotic genome of approximately 400 Mb, SynGears Omni successfully achieved full-process automation from raw sequencing data to downstream analysis, completing in two days what traditionally takes about two weeks of manual work.

The team clearly remembers the moment they finished running the process and got the test results: "This completely exceeded our expectations."

Currently, SynBioLab's computing team has achieved full AI collaboration.

Four Major Engines: Beyond Analysis

Notably, omics capabilities are not the endpoint of SynGears Omni, but rather its important entry point for discovering new enzymes, connecting protein design, and empowering biomanufacturing. The team plans to further expand to metabolomics, metagenomics, pangenomics, single-cell omics analysis, and other capabilities, gradually forming a complete intelligent R&D platform covering "sequencing-analysis-design-verification."

SynGears Omni also possesses significant advantages in four aspects: decision collaboration, information integration, industrial application, and data security, particularly suited for R&D teams to meet full-chain needs of omics plus enzymology.

① Decision Collaboration: Five-Agent Cluster, Decision Tree Precision Scheduling

SynGears Omni operates through 5 sub-agents working collaboratively, supporting natural language interaction, full-process task monitoring, and integrated visualization interface, with logs/versions/parameters/DAG diagrams fully saved, achieving 100% reproducible results, delivering a fully managed experience similar to an exclusive bioinformatics expert team.

The platform innovatively adopts decision tree intelligent scheduling, which can automatically select analysis strategies based on species and data types, with anomaly monitoring and dynamic adjustment, effectively reducing model hallucination risks. Additionally, the platform possesses self-evolution capabilities, can accumulate experience for continuous optimization, better fitting research scenarios.

"SynGears Omni can read user preferences during usage, recommend better solutions based on these preferences, and the agent can record this information to Memory and recommend better parameters," SynBioLab's computing team explained with an example when introducing the platform's self-evolution capabilities.

Decision tree design enables users to complete full-chain tasks from omics data parsing to enzyme source discovery on SynGears Omni without relying on any external tools or switching between multiple workflows.

② Information Integration: Self-Built Data Foundation, Building Continuously Growing Knowledge Graph

Facing the common industry dilemma of scattered data and restricted access, SynBioLab innovatively created a self-operating exclusive biological data platform—it is not a simple information warehouse, but a continuously growing, self-updating digital life graph. By smelting species genome fragments, evolutionary clues, and annotation information scattered globally into an internal network covering plants, animals, fungi, and thousands of life forms, it provides underlying data support for multi-omics analysis.

Traditional processes often rely on "borrowing routes" overseas, easily affected by network fluctuations, permission restrictions, and information fragmentation. SynGears Omni, relying on this localized information hub, achieves stable data scheduling and rapid response.

Notably, this data platform synchronizes and evolves weekly with three major global databases, truly achieving "zero delay, fully autonomous" data acquisition.

Industrial Application: From Genome to High-Value Molecules, End-to-End Discovery

In practical applications, SynGears Omni performs excellently in the biomanufacturing field, especially in research on various high-value functional molecules (such as natural product active molecules). It can directionally mine target product synthesis pathways and key enzymes from multi-omics data, and conduct molecular docking and virtual screening combined with target product molecular characteristics, effectively reducing false positive rates.

SynBioLab's computing team expressed deep feelings about this in the interview: "Taking natural products as an example, one of the difficulties is enzyme redundancy—one enzyme may catalyze multiple reactions, while multiple enzymes may also catalyze the same reaction. Facing such unknown pathways, SynGears Omni can complete genomics enzyme mining work with one click, and subsequent steps only require predicting function through protein structure and interaction analysis and conducting experimental verification. It can be said that SynGears Omni is gradually bridging the gap between enzymology and omics, making the two combine more closely."

Additionally, the intelligent agent can automatically judge and analyze model-predicted enzyme-substrate candidate pairs, screening out more reference-worthy results, providing intuitive and usable operational objects for laboratory wet experiment verification, significantly accelerating the target product-related enzyme screening and pathway optimization process.

④ Data Security: Encapsulated in a Box, Plug and Play

Meanwhile, SynGears Omni supports local deployment and containerized operation, flexibly adapting to university, research institute, and enterprise server environments. All data remains running and stored within the user's local server—not uploaded, not transmitted externally—while ensuring data security and privacy compliance, it is also more suitable for research scenarios involving core strains, proprietary sequences, and industrial biological data.

SynBioLab's computing team further supplemented the convenience brought by encapsulation design: "First, complex genomes and large genomes can be easily migrated between different servers—the encapsulated platform is like a portable box with its own environment, running directly without reconfiguration. Second, encapsulation greatly lowers the usage threshold. Without encapsulation, users must manually install all dependencies for each server, a cumbersome process; after encapsulation, the platform comes with its own runtime environment, plug and play on any server, significantly improving user experience."

From Gene to Intelligent Manufacturing: The Technical Foundation of SynGears Omni

The omics capabilities and agent architecture upgrades carried by SynGears Omni did not come out of thin air, but are built upon SynBioLab's long-accumulated "AI + synthetic biology" technology system.

As one of the earliest domestic teams to introduce AI into synthetic biology research, the company has accumulated rich practical experience in AI-driven rational enzyme protein design. (See: From Understanding a Tree on the Plateau to Completing the Ten-Thousand-Ton Path: SynBioLab's AI Biomanufacturing Practice)

Relying on this platform, the team successfully developed the BioFlux dual-module automated workflow, integrating multiple algorithms to achieve full-chain automation from high-throughput protein screening to high-precision structural analysis, screening over 3 million ligand-receptor combinations daily on average, completing screening tasks that would take traditional wet experiments months.

On this basis, SynBioLab's computing team, through AI-driven rational design and machine learning-assisted directed evolution strategies, has successfully developed a series of high-performance enzyme mutants and completed industrial transformation, and has also achieved solid results in the field of genome mining.

These capabilities were initially consolidated into Xuanzhu Computing's core protein computing platform SynGears. Now, the team has further integrated years of accumulated omics analysis experience, natural product research methodology, and agent technology into it, upgrading to form SynGears Omni.

Through natural language interaction, multi-agent collaboration, and automated bioinformatics analysis, SynGears Omni is gradually transforming large-scale omics research processes that were previously highly dependent on expert experience into reusable, scalable intelligent computing capabilities.

Just as SynBioLab's team expressed their belief in the interview: The complexity of biology is not an obstacle for AI, but precisely where AI should shine brightest.

"What SynGears Omni opens up is not just computational tools, but the analysis logic and domain knowledge we've accumulated over years in green and efficient synthesis research of high-value-added functional molecules. We hope it can become infrastructure for more researchers in the industry, jointly accelerating the development process of the natural products treasure trove.

We also sincerely welcome researchers and partners with shared visions to work with us to improve this platform—every use and feedback from users is nourishment that makes SynGears Omni smarter."

It is understood that SynBioLab will further improve SynGears Omni's multi-sub-agent collaborative architecture, advance containerization and local APP deployment, ensure data security, and continuously upgrade the database, focusing on building China's leading end-to-end AI agent in the biomanufacturing field.

In Closing

SynGears Omni is currently in the system internal testing phase and has not yet been opened for public access. It is expected to officially open for public beta testing in the second half of 2026. If your research involves scenarios such as non-model organism genome assembly and annotation, natural product synthesis pathway mining and key enzyme discovery, engineered strain system optimization and metabolic pathway design, multi-omics data joint analysis and intelligent reasoning, welcome to continue following!

—The End—