Development preview. Not validated. Not for clinical decision making.
Diffuse shading marks a side chain implicated only by fingerprint similarity, which scores the side chain as a whole and does not identify specific atoms. Solid circles mark atoms matched by substructure or MCS.
Please select an antibiotic.
PenicillinX predicts which beta-lactam antibiotics are likely to cross-react by measuring the chemical similarity of their side chains.
Every drug here shares the same beta-lactam core, so the core cannot explain why a patient reacts to one drug and tolerates another. Hypersensitivity is instead largely driven by the side chains hanging off that core: the R1 group, and for cephalosporins the R2 group. Each molecule is drawn with its beta-lactam ring shaded blue so you can see the shared anatomy, and matched side chain atoms circled in red (R1) or blue (R2).
No. This uses established, deterministic chemoinformatic algorithms that analyse and compare the chemical and topographical properties of the target molecular pairs according to pre-determined criteria and similarity thresholds. The advantage of this method is transparency and reproducibility as compared to a machine learning model.
Compare answers "my patient reacted to these drugs, what do they share?". Pick up to three antibiotics and see them drawn side by side, with the prediction for every pair underneath.
Matrix shows all 51 antibiotics against each other at once. Cell colour based on Hutten et al. (2025); a black dot is this model's prediction, a blue dot is a prediction based on R2.
| No cross-reactivity | |
| Cross-reactivity | |
| Discrepancy between source studies | |
| No data | |
| Same drug |
| R1-driven | |
| R2-exclusive - R1 similarity fell below threshold | |
| R1 and R2 both flagged independently |
The R2-exclusive case is the novel finding of the accompanying work: pairs that look unrelated by their R1 groups, but share an R2 group closely enough to matter.
The model incorporates seven different chemoinformatics algorithms testing molecular similarity. A pair is predicted to cross-react if any of them flag positive. Four compare whole-molecule fingerprints and return a single similarity score; three identify the specific atoms responsible and can therefore drive the highlighting. Where only fingerprint methods are positive, the implicated side chain is shaded as a whole instead, since no individual atoms are identified.
Results must not inform clinical decision making to any extent.
Original matrix derived from:Hutten, E.M., Bulatović Ćalasan, M., Trubiano, J.A., Pleijhuis, R.G.-J. and Terreehorst, I. (2025), Discrepancies in Beta-Lactam Antibiotics Cross-Reactivity: Implications for Clinical Practice. Allergy, 80: 3108-3114.