MetaFLUX by All Saints | BioHack 2026 | PacificoBio

Introslide

Inspiration

Fungal strain optimisation is still too slow, expensive, and manual. In many cases, testing a single strain under one growth condition can take up to two weeks, and repeating that across multiple media conditions quickly becomes costly and low-throughput. We were inspired by the idea that genome data should be enough to get an early, useful prediction of how an organism will grow before committing to wet-lab experiments. MetaFLUX was built to bridge that gap by turning fungal genome data into fast metabolic insight.

What it does

MetaFLUX is a computational tool for fungal metabolic analysis. A user uploads a fungal genome file, the platform extracts the encoded protein information, builds a genome-scale metabolic model, and simulates growth computationally. From that, MetaFLUX predicts biomass production, growth rate, and nutrient requirements, and returns an intuitive summary of the organism’s metabolic potential and likely optimal growth medium. This helps researchers compare strains and media conditions in minutes rather than relying only on slow trial-and-error experiments.

How we built it

We built MetaFLUX using Python for the scientific backend and JavaScript for the user-facing interface.

On the backend, our workflow starts with a fungal genome input and focuses on the protein FASTA derived from that genome. We then use CarveMe or ModelSEED to reconstruct a genome-scale metabolic model (GEM) from the protein annotations. Our approach is grounded in GEM reconstruction concepts.

Once the GEM is generated, we analyse it in Python using COBRApy, which allows us to run constraint-based metabolic simulations. This is where we estimate outputs such as biomass production, growth rate, and the effect of available nutrients on predicted growth. On the frontend, we designed a simple JavaScript-based upload workflow so the user can drag and drop a fungal genome and receive an interpretable result rather than raw model output.

Challenges we ran into

One challenge was turning a complex systems biology workflow into something simple enough for a hackathon demo and understandable to a broad audience. Genome-scale metabolic modelling involves multiple steps, and each step has its own assumptions and technical requirements.

Another challenge was connecting the pipeline end-to-end: starting from genome-derived protein data, generating a usable GEM, and then making sure the model could be analysed reliably for growth and biomass outputs. We also had to think carefully about how to present the results in a way that feels useful to a non-specialist user, rather than overwhelming them with metabolic network details.

Accomplishments that we're proud of

We are proud that we built a working concept that takes a fungal genome and turns it into something biologically meaningful and actionable. Instead of stopping at model construction, we connected the full pipeline through to simulation and interpretable output.

We are also proud that MetaFLUX does not just describe metabolism in theory. It aims to answer practical questions researchers actually care about: How well might this strain grow? What medium conditions look promising? How can we reduce unnecessary experimental testing? Building a pipeline that is both technically grounded and easy to explain was a big win for us.

What we learned

We learned how powerful genome-scale metabolic models can be when combined with accessible software design. A genome is not just a static sequence file, it can be turned into a predictive model of biological behaviour.

We also learned that there is a real need for tools that sit between raw omics data and experimental decision-making. From a technical side, we gained experience in integrating Python-based modelling tools like CarveMe and COBRApy with a more approachable JavaScript frontend. From a project side, we learned how important it is to simplify complex biology without losing the science.

What's next for MetaFLUX by All Saints | BioHack 2026 | PacificoBio

Our next step is to make MetaFLUX broader, more robust, and more useful in real biological workflows. Right now we are focused on fungi, but the same framework could be extended to other organisms with metabolic data, including bacteria, yeasts, plants, and even mammalian cells. We also want to expand beyond biomass and growth rate so the platform can optimise a wider range of environmental and media conditions.

Longer term, we see MetaFLUX becoming a decision-support tool for bioprocess design: helping researchers optimise culture conditions, prioritise experiments, and design smarter, more efficient bioreactor workflows from genome data alone.