Go beyond lab as code with CI CD and tech best practices

Post by
Go beyond lab as code with CI CD and tech best practices

In the realm of biotech, where wet labs are the heartbeat of discovery, seamlessly integrating computational elements is the key to accelerated innovation. A recent thought-provoking webinar by Scispot shed light on this integration through the lens of 'Lab as Code'. This concept aims to programmatize lab workflows, bringing a new level of efficiency and precision to the fore.

The webinar panel featured Scispot's co-founder, Satya Singh, Jesse Johnson of Merelogic, and John Whittaker from Invitae, each bringing a unique perspective to the table. They shared valuable insights on how transitioning to a code-driven lab workflow could be a game-changer in managing lab operations.

Scispot has cutting edge data infrastructure and analytics that are built to streamline biotech R&D

Bridging the current gaps

The discussion kicked off with an outline of the existing challenges. Labs today juggle multiple systems, each with its own set of rules, leading to a fragmented workflow. Manual methods are still in play due to a lack of modularity in existing automation systems. Moreover, the absence of version control, a system to track changes in code, leaves room for errors and inefficiencies.

The three challenges highlighted above hinders the lab efficiency as the lab's data infrastructure scales.

  1. Modularity & Bespoke Experimentation: Even the most advanced lab automation systems and architecture struggle to provide the kind of modularity that bespoke experiments demand, while simultaneously being powerful enough to function as production lines. Manual methods, with all their inherent inefficiencies, still remain the most versatile tool in a scientist's arsenal.
  2. Complexity & Testing Environments: As lab automation becomes more complex, the reliability of testing environments decreases. This becomes particularly evident with larger integrations, and lab orchestration. The concept of the Digital Twin as a tool for simulations is still relatively foreign to the industry, and testing on production equipment is almost always required.
  3. Version Control & Data Capture: A significant pain point in today's labs is the absence of effective version control and comprehensive data capture mechanisms. Industry standard low-code software almost always requires internal libraries and plugins to manage run data and communicate with external services. The interdependence of software and hardware components, coupled with the integration of third-party services makes versioning an uphill battle. 

Envisioning a Code-driven lab

John Whittaker, who leads lab automation at Invitae painted a picture of a future where a single codebase defines the entire lab process, eliminating redundant systems. This unified codebase promotes easier management, updates, and version control. He also touched on the idea of automating code deployment, ensuring the right code is in action at the right time, thus minimizing errors.

Scispot's Data Lake Infrastructure

Scispot, a leading platform, offers powerful solutions for building robust data infrastructure in the biotech industry. By bringing disparate data sources together and providing rationalization capabilities, Scispot simplifies data management and enhances data integrity. Its orchestration engine enables the creation of knowledge graphs and facilitates metadata management. Additionally, Scispot ensures that R&D data is readily available and compatible with machine learning pipelines, eliminating the need for extensive schema updates and devops work.

John's vision is to leverage proven, scalable best practices and tools from successful technology companies to modernize the lab automation software stack, driving scientific discoveries and process innovation. To achieve the vision, we need the following modules to come together in a product. Scispot has already built some of these modules, and committed to enhancing lab automation features in the upcoming releases.

  1. Worfklow Designer: Define entire lab processes as code
  2. Lab Controller: Automate the versioning of that code, and capture all dependencies
  3. Lab Controller plus Agents: Capture telemetry data from all instrumentation and software
  4. Workflow Designer along with Scheduler & Agents: Retire outdated desktop scheduling software for a cloud native scheduler (Workflow Designer + Scheduler + Agents) that can integrate with industry standard hardware
  5. Connectors: Build the scheduler such that it seamlessly connects to external LIMS

Scispot is the first LIMS system that is thinking of LIMS as not just a scalable data warehouse, but also as a connector to orchestrate various lab workflows so you can run your lab as code. Here is an example of how these features complements the vision of running your lab as a continuous integration and continuous deployment pipeline.

Concrete steps towards unified workflows

The 'Lab as Code' model proposes replacing disparate desktop scheduling software with a centralized, cloud-native scheduler. This scheduler would be designed with testing in mind, ensuring robust communication between different lab systems and a smoother workflow. 

Moreover, a significant part of the discussion was dedicated to creating better connectors. These connectors would bridge the gap between Lab Information Management Systems (LIMS) and scheduling software, facilitating a seamless flow of information.

Our data lake infrastructure surpasses other traditional eln and lims systems, creating an all encompassing environment for research

Community engagement & learning from tech

The webinar didn’t just stop at discussing the concept. It opened the floor to the community, encouraging attendees to share their ideas and questions. This collaborative approach is seen as a vital step towards refining and realizing the 'Lab as Code' vision.

Additionally, the panelists highlighted the potential of learning from the tech industry, which has made strides in similar areas through Continuous Integration and Continuous Deployment (CI/CD) pipelines.

Jesse, who runs a data consulting company for bio called Merelogic focussed on breaking down the barriers between wet lab and computational persona.

Jesse highlighted the following challenges that he has witnessed in a lab with both wet and dry workflows:

  1. Process details versus Static summary: Lab teams need to think in terms of the context-specific steps needed to deal with corner cases and make the biology. To properly interpret instrument data, digital teams need a static view of specific details hidden in the complex process.
  2. Flexibility versus Consistency: The bespoke nature of experiments that makes automation software so difficult also makes building repeatable analysis or cross-experiment analysis much harder.
  3. "Throw it over the wall" mentality: Biologists are used to planning experiments without input from data teams. Data scientists know what data they need, but often don't feel empowered to contribute to planning.

Lab as code philosophy can break down these barriers, and help both wet lab and computational biologists to collaborate in lab. This requires implementing a process around your lab data, people, and tools. Some of the examples are: 

  1. Collaborative planning between wet lab and digital teams
  2. Detailed experiment definition before the experiment starts
  3. Capture process view in a translatable digital format
  4. Automated merging of experiment details and readout

Scispot features to help labs go beyond traditional LIMS systems to run lab as code

Scispot has built the world's first LIMS with API-first mentality. Everything within the platform is accesssible with a secure API end point. This helps to run and connect lab workflows programmatically.

Scispot's toolkit has the following modules that help facilitate the lab as code philosophy:

  1. GLUE: The connectors and agents to bring disparate data from instruments and robots together in a secure cloud
  2. Wet Lab tools: Modern version of ELN and LIMS features for wet lab scientists that is chemically and molecular biology aware
  3. Computational tools: Embedded Jupyter Hub and RStudio for computational biologists to easily connect wet lab with computational workflows
  4. Search & Discovery: Easily find your data using AI co-pilot

Sneek Peek into Scispot's roadmap specific to lab as code

Scispot is working on a future vision that not only focusses on lab as code but lab as English language that empowers wet lab scientists to easily write their queries and python scripts without writing a single line of code. If you'd like to learn more, you can schedule a call directly with our co-founder here


'Lab as Code' isn’t a fleeting trend but a solid framework that could redefine lab operations in the biotech industry. By fostering a symbiotic relationship between wet labs and computational elements, and with a nudge from collaborative community engagements, the biotech domain is on the brink of embracing a more streamlined, effective operational model.

The 'Lab as Code' paradigm is not only about modernization but about harmonizing the rhythm of lab operations to the tune of technological advancements. Through this lens, the future of lab automation in biotech appears not just promising, but well within reach.

Sign up for Scispot

What’s a Rich Text element?

The rich text element allows you to create and format headings, paragraphs, blockquotes, images, and video all in one place instead of having to add and format them individually. Just double-click and easily create content.

Static and dynamic content editing

A rich text element can be used with static or dynamic content. For static content, just drop it into any page and begin editing. For dynamic content, add a rich text field to any collection and then connect a rich text element to that field in the settings panel. Voila!

How to customize formatting for each rich text

Headings, paragraphs, blockquotes, figures, images, and figure captions can all be styled after a class is added to the rich text element using the "When inside of" nested selector system.