Answers are grounded in this paper and cite their sources.
Antifungal susceptibility tests report how much a drug slows a population, not how that slowdown is distributed among its cells. A culture can lose half its activity because every cell runs at half speed, or because half the cells stop while the rest continue almost unaffected. We distinguished these cases in Candida using a public dataset of 6,227 single-cell Raman spectra from seven strains of C. albicans, C. glabrata andC.krusei,grown inheavy water(D2O) and exposed toamphotericinB,voriconazole, micafungin or flucytosine. The deuterium signal, expressed as C–D/(C–D+C–H), separated metabolically active from inactive cells with almost no overlap, which made the shape of each single-cell distribution directlytestable.Wefittedgraded,all-or-noneandhybridmixturemodelsandestimatedthesilent fraction withbootstrap intervalsthatalsoresampled thereference populations.The twoC. albicans strains that were partially inhibited responded in a graded way in all seven conditions. In all three nonalbicans strains, 11 of 14 partially inhibiting conditions contained a silent subpopulation, with all four drugs;C. krusei reached 58% silent cells (95% CI 48–65%) under micafungin at 1 mg/L. At the same population-mean inhibition of about 40%, C. albicans had no silent cells while non-albicans cultures had 18–35%. Silent cells gave normal fingerprint and C–H signals, so they were intact cells rather than poor spectra. Bulk readouts cannot see this split; single-cell D2O Raman resolves it within six hours.
No takes yet. Share an insight, caveat, or question.
Seyedeh Rahemeh Amiri (2026) studied this question.
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: