AURA: Active-Response Attribution under Treatment Ambiguity in Bacterial Cytological Profiling

2026-06-15

AI Summary

This study addresses the challenge in bacterial morphological analysis where not all administered antibiotics in a combination regimen necessarily exert biological effects, necessitating accurate inference of the active subset from observed cellular phenotypes. The authors formulate this as an energy-based inverse attribution task with subset constraints, decomposing morphological residuals and identifying the combination of antibiotic response atoms that yields the lowest reconstruction energy—thereby enabling label-free identification of effective drugs. The proposed method incorporates inductive bias through explicit subset constraints and an evidence-aware abstention mechanism, substantially enhancing generalization. Evaluated on cross-replicate experiments with *Escherichia coli*, the approach achieves a 95.47% exact-match accuracy in recovering the true set of active antibiotics.

Abstract

When a bacterial sample is exposed to several antibiotics, not every applied drug necessarily acts: if the organism is resistant to one of them, that drug leaves no morphological trace. The clinically meaningful quantity is therefore not which antibiotics were applied, but which ones were active. We show that these two are sharply decoupled in real E. coli microscopy - naively assuming the applied combination equals the active one is correct only about 37% of the time - yet existing computational tools are ill-suited to recovering the active set. Forward perturbation models such as scGen, CPA, and IMPA are designed to predict appearance from treatment, not the reverse, and inverting them degrades sharply; discriminative image classifiers tend to memorise strain- and batch-specific texture and fail to transfer across experimental replicates. We introduce AURA, which reframes the task as constrained, energy-based inverse attribution. Its central inductive bias is that the active set must be a subset of the applied set; this collapses the candidate space and lets AURA infer the active subset of applied antibiotics by decomposing residual morphology into antibiotic response atoms and selecting the subset with the lowest reconstruction energy, using no strain label at test time. AURA-E adds evidence-aware abstention, withholding a prediction when candidate explanations remain near-equally plausible. On cross-replicate transfer in an E. coli cytological profiling dataset, AURA recovers the active antibiotic combination with 95.47% exact-match accuracy.

Problem

• antibiotic activity
• treatment ambiguity
• bacterial cytological profiling
• active-response attribution
• morphological response

Innovation

• inverse attribution
• energy-based model
• treatment ambiguity
• bacterial cytological profiling
• active-response inference