Phenomics - the great missing omics

There is a major, gaping hole at the center of biology.

Molecular biology has been digitized. You can get terabytes of standardized, high quality data from a single genomics/transcriptomics/proteomics run.

But, what we ultimately care about is not the molecules - but the impact the molecular change had on the functioning of the whole organism.

Yet when it comes to measuring health & phenotypes, the state of the art is manual, sometimes semiquantitative measurements with data written in paper notebooks.

It’s a 10,000x disparity in generating data from molecules vs whole organisms - even in the most advanced labs in the world.

There isn’t even a commonly used word for “large-scale, standardized, organism-scale phenotypic data generation”.

That’s insane. And it’s a major contributor to why translation fails so often.

In short, “Phenomics” is a great missing omics.

So - building a “phenomics” device was our first goal at Olden. 

That meant:

1. Solving the multi-animal tracking problem - maintaining correct IDs accurately, over long periods of time in complex environments.

2. Miniaturizing the hardware 

3. Achieving a low cost point that would beat manual analysis.

The first stage of the work is now complete and published, here.  

Key results:

-Tech: Smart Lids - phenomics devices that continuously monitor all animals in their cage, gathering data on health & behavior 10 times per second, 24/7.

-Data: 21 metrics across activity, motion, eating, drinking, fighting, rearing, climbing, sleeping, social behaviors and positioning. 

-Segmentation accuracy: 99.5% concordance with human

-Multi-animal tracking accuracy: 97% average (even at 5 mice)

-Metrics precision: 82-99% (some room for improvement but achieved at very low computational cost)

-Robustness: works in messy, real-world scenarios, robust across occlusions, coat colors, jumps, different lighting, beddings, nesting materials, enrichments and 5 major cage types

There is room for growth: improving precision for certain metrics, expanding to new cage types and species, and adding contact-based metrics like weight and cognition.

But - the key is - the system wasn’t built for an academic publication, but to be an accessible, everyday tool from day 1.

We now have hundreds of scientists across 5 continents using the tech, making discoveries daily. The first of our users’ papers will be out soon.

So - if you have a friend or colleague doing animal research, send them this.