Faster Bioinformatics with Seqera AI

Today, we’re excited to announce Seqera AI – a bioinformatics agent purpose-built for the scientific lifecycle. We’ve engineered Seqera AI to meet scientists at every stage of their work –

✔ Suggesting tested and validated pipelines, modules, and sub workflows from the Nextflow community
✔ Generating Nextflow code fully aligned with best practice for early stage development
✔ Testing, validating, and self-correcting the pipeline code it generates and providing seamless nf-test support
✔ Asking bioinformatics questions and receiving contextually relevant answers that inform your analyses.

LLMs have come at a critical moment – one where we are continually thinking about how we can lower the barriers to entry for bioinformatics and enable everyone in science to leverage code for more complex analysis. While LLMs have taken us beyond where we could have imagined a few years ago, the high complexity and context required for bioinformatics can be lost in translation, potentially leaving scientists feeling left behind.

How Seqera AI is built to support Bioinformatics

What does having a better AI experience for science truly mean? At Seqera, we think it comes down three key things:

1. Purpose-built – Built to enable bioinformaticians with Nextflow development best practices. Seqera AI helps ensure that bioinformaticians can get targeted, comprehensive help regardless of their experience level.
2. Comprehensive – Beyond just a chat interface, Seqera AI gives researchers the ability to test their code in the interface so they can be more sure that it will run.
3. Nextflow and nf-core aware – Programmed with a deep understanding of Nextflow, common bioinformatics tools, and the overarching scientific community.

We’ve crafted Seqera AI with these principles in mind, and after extensive testing with scientists – both across Seqera and in our community – we’re confident that we’ve built a tool that significantly enhances the existing AI experience.

Getting closer to the right code from your first message

Whether you’re generating net new code based on a prompt, extending existing code with a new module or code from another pipeline, or converting an existing pipeline into Nextflow, Seqera AI uses built-in Nextflow awareness to get pipeline code correct on the first try.

Better debugging and troubleshooting

By building on community best practices Seqera AI has deep knowledge of the errors that you can encounter when writing and running Nextflow and is able to identify the root cause of errors, help troubleshoot, and suggest edits to code to unblock your development.

Automatic self-correcting

In building Seqera AI, we’ve given it more than the ability to read and write code. We’ve natively built the ability to pair with bioinformatics test data and generate local test scripts for your code or automatically generate and run unit tests. This enables an automatic feedback loop for Seqera AI, where you not only get better generated code from the start but also code that is pre-tested and validated.

When you start to put these pieces together they get incredibly powerful. For instance, if you want to convert a pipeline from Bash/CWL/WDL to Nextflow, Seqera AI can not only give you the initial conversion, but also run the stages of the code that it generates with sample data and iteratively correct any code that yields runtime errors, let’s take a look –

Native Interpretation of Results

Beyond answering questions and writing code, we believe AI can be a powerful tool for helping scientists dig into results and more quickly identify interesting patterns. However, just like we’ve needed to bring the right context for writing and testing pipeline code, it’s been key to figure out how we can get the right context on your pipeline results into Seqera AI.

So today we’re excited to announce that we’ve created a native integration with MultiQC where you can enable automatic, in-line analysis of MultiQC reports by Seqera AI.

While this isn’t fully available yet, it will be released soon as an opt-in configuration in MultiQC. Even more exciting, we’ve built this as a fully extensible endpoint in Seqera AI, so that any bioinformatics tool can build their own AI integration.

As with everything we approach, building alongside the community is a key part of how we move forward, so we encourage you to try Seqera AI and let us know your ideas for how we can improve – we can’t wait to see what you use Seqera AI to build!