Meta-Flux is collaborating with Celtic Biotech, a fellow Dublin company developing snake-venom-derived cancer therapeutics, to build an in silico picture of where its lead candidate is likely to work. The collaboration, supported by the European Innovation Council, aims to map where a snake-venom-derived cancer therapeutic works and where it does not.

The aim is narrow and practical. Rather than describe a molecule in the abstract, the innovators are applying advanced AI to study CB-24 against real tumour biology, characterising the patterns of sensitivity and resistance that separate the patients most likely to respond from those unlikely to benefit.

Celtic Biotech develops cancer therapeutics derived from snake venom, a source with a long and under-exploited pharmacological history.

Its lead candidate, CB-24 (crotoxin, isolated from the venom of the South American rattlesnake), is directed at solid tumours where existing options are limited and where the need for a differentiated mechanism is most acute.

Candidates of this kind rarely act uniformly across a tumour type. The clinically useful question is not whether CB-24 is active, but in which tumour contexts, against which molecular backgrounds, and for which patients its activity is durable. That is a systems-biology question, and it is where the modelling begins.

Together with Celtic Biotech, Meta-Flux is applying artificial intelligence to study CB-24 across 831 oncology cell lines, each one placed alongside harmonised multi-omic and clinical data so that response can be read against the underlying biology rather than treated as an isolated readout. From this, the teams model the patterns of sensitivity and resistance across the panel, tracing observed response back to the molecular features that appear to drive it.

Because every prediction is anchored to a mechanistic model rather than a statistical correlation alone, the output is not a ranked list of responsive lines but a mechanistic account of why CB-24 works where it works. That is the part that transfers from cell lines to patients, and that supports decisions once the candidate moves towards the clinic.

The value of this work is in what it feeds downstream. A mechanistic map of sensitivity and resistance informs patient stratification, giving Celtic Biotech a defensible basis for selecting the populations in which CB-24 is most likely to succeed. The same map supports clinical development, from trial design and enrolment criteria to biomarker strategy, and gives regulatory planning an evidence trail that connects the candidate’s activity to specific biology.

04.08.2026.