Our Platform

Our 3D Process

Our product candidates stem from the Axcella Design Platform, which combines deep metabolism expertise, correlative reasoning algorithms, proprietary non-clinical systems, advanced analytics, and an expansive safety database.


Defining Our 3D’s


The body is replete with a vast amount of EMMs, which help to regulate many pathways and functions. Our development process enables us to identify product candidates with multifactorial potential from a vast landscape consisting of more than 30 million possible EMM combinations. We begin this process by identifying multiple biologies and pathways that have the potential to be regulated or signaled by EMMs. We then consider whether such an EMM intervention may have a meaningfully beneficial impact on disease and/or may support healthy structures and functions within the body.

Using the world’s scientific and medical literature, public databases and our own metabolic profiling from biorepository samples, we have characterized approximately 75 potential applications for EMM compositions.


We have built a proprietary knowledge base, called the AxcellaKB, that consists of:

  • Extensive and ever-expanding data from clinical trials and published literature on the nonclinical and clinical use of individual EMMs and simple EMM combinations
  • A comprehensive data collection of metabolic profiles from human biological samples
  • EMM dosing and safety precedents
  • Non-clinical proprietary models such as human primary cell systems
  • Omics-characterization experiments (i.e. RNASeq, metabolomics)
  • Pharmacokinetics of EMMs alone and in combination
  • Advanced analytical approaches such as machine learning to integrate these data sources.


The output of our extensive data science and systems biology process are EMM compositions (product candidates) with the potential to treat complex diseases. These candidates then enter clinical-stage development. Axcella’s lead product candidate (AXA1125) has advanced from initial concept to Phase 2 development in approximately four years, cutting the standard drug development timeline roughly in half.

Learn more about our pipeline