Interlaboratory benchmarking of differential scanning fluorimetry reveals sources of variability in protein thermal stability parameters

Brule, S.; Chevreuil, M.; Burastero, O.; Abian, O.; Anderluh, G.; Aumont-Nicaise, M.; Bagdonas, M.; Birchenough, H.; Buell, A. K.; Charnavets, T.; Crousilles, A.; Danilenko, N.; Epe, M.; Fabini, E.; Feoli, A.; Garcia-Alai, M.; Gasper, R.; Griss Pascualli, E.; Houser, J.; Jowitt, T.; Kuntzmann, A.; Langer, A.; Lapouge, K.; Lau, K.; Le Roy, A.; Lopez-Mendez, B.; Malempre, R.; Mas, C.; Matagne, A.; Matulis, D.; Mueller, A. H.; Pasqualato, S.; Paulenova, E.; Sbardella, G.; Streicher, W.; Struve Garcia, A.; Tsvetkov, P. O.; Uebel, S.; Vandevenne, M.; Vega, S.; Velazquez-Campoy, A.; Walton, R. W.; Z

doi:10.64898/2026.10.12.758416

Home/Archive/bioRxiv

bioRxiv · 10.64898/2026.10.12.758416

Interlaboratory benchmarking of differential scanning fluorimetry reveals sources of variability in protein thermal stability parameters

View original source ↗Export metadata ↓

Abstract

Differential scanning fluorimetry (DSF) is widely used to assess protein thermal stability and to compare protein variants, ligands, or formulations with limited sample consumption. Yet, the reproducibility of DSF derived parameters across laboratories, instruments and analysis workflows remains poorly defined. Here, we report a two phase interlaboratory benchmark using a shared VHH sample across 26 laboratories, 28 instruments and seven DSF instrument models. In Phase 1, participants applied their routine acquisition and analysis workflows, whereas in Phase 2, acquisition conditions were harmonised. Datasets were analysed locally and then centrally reprocessed with MoltenProt. Participant reported apparent melting temperatures (Tm) showed good within instrument repeatability but moderate inter instrument dispersion. Harmonisation of acquisition conditions did not substantially reduce this dispersion. Conversely, centralised reprocessing narrowed the distribution among retained datasets, indicating contributions from analysis workflows and data acceptance criteria in addition to instrument dependent factors. Comparison with differential scanning calorimetry (DSC) further showed that apparent Tm depended strongly on the fluorescence readout. For this VHH, single wavelength fluorescence at 350 nm showed the closest agreement with the DSC transition temperature, whereas the 350/330 nm ratio and barycentric mean generally yielded higher values. Furthermore, DSF derived apparent unfolding enthalpies were more variable than apparent Tm. Together, these findings show that DSF can be highly repeatable within a defined instrument and workflow, but that quantitative comparability across instruments depends on instrument specific effects, the choice of fluorescence readout and analysis strategy. Improved comparability will require transparent reporting of acquisition and analysis conditions, explicit data quality criteria, and well characterised reference materials to monitor instrument dependent offsets.

Explore related subjects

biophysics Preprint

Keep this discovery

Explore connections, maps & timelines

BibTeX RIS

Brule, S., Chevreuil, M., Burastero, O., Abian, O., Anderluh, G., Aumont-Nicaise, M., Bagdonas, M., Birchenough, H., Buell, A. K., Charnavets, T., Crousilles, A., Danilenko, N., Epe, M., Fabini, E., Feoli, A., Garcia-Alai, M., Gasper, R., Griss Pascualli, E., Houser, J., Jowitt, T., Kuntzmann, A., Langer, A., Lapouge, K., Lau, K., Le Roy, A., Lopez-Mendez, B., Malempre, R., Mas, C., Matagne, A., Matulis, D., Mueller, A. H., Pasqualato, S., Paulenova, E., Sbardella, G., Streicher, W., Struve Garcia, A., Tsvetkov, P. O., Uebel, S., Vandevenne, M., Vega, S., Velazquez-Campoy, A., Walton, R. W., Z. 2026-10-13. Interlaboratory benchmarking of differential scanning fluorimetry reveals sources of variability in protein thermal stability parameters. https://doi.org/10.64898/2026.10.12.758416

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related preprints

bioRxiv2026

Structural order emerges without mechanical order inside living cells

Intracellular organization is vital for cell function, and disruption of organelle positioning is connected to severe diseases, such as cancer. Given the small length scale of cells and the importance of Brownian motion, maintaining intracellular organization requires mechanical interactions to remain stable. Here, we test the common view that intracellular organization is reflected in intracellular mechanical properties. Using the recently introduced active mechanical fingerprint, we quantify intracellular mechanical properties and active forces to create mechanical maps of patterned RPE1 cells. The organization of actin, microtubules, the Golgi apparatus, the endoplasmic reticulum, and the nucleus reveals pronounced structural polarization across the cell population. To overcome limitations of simple averages, we use a Wasserstein barycenter to calculate an optimal intracellular organization that explains all measurements with minimal cellular distortion. To our surprise, and in contrast to previous reports, we find neither a statistically significant spatial mechanical pattern nor any relevant correlation between the tested structures and intracellular mechanical properties. While there is clear local mechanical anisotropy, the large heterogeneity of mechanical values outweighs any intracellular mechanical gradient. This is further supported by large cell-to-cell variability. Thus, despite having the same genome, cells function properly while their intracellular mechanical properties vary over two orders of magnitude.

biophysics↗

bioRxiv2026

CALISTO: A Standardized and Open-Source Software for Magnetic Tweezer Force Calibration

Magnetic tweezers (MT) enable multiplexed single-molecule force spectroscopy while applying controlled forces and torques to biomolecules. They resolve forces below 1 pN and maintain constant forces over extended periods, but accurate calibration remains challenging due to drift, tracking artifacts, camera noise, and bead variability. Here, we present CALISTO (CALIbration Suite for magnetic Tweezer Operation), an open-source software for MT force calibration. CALISTO analyzes position traces of DNA-tethered magnetic beads and surface-bound reference beads for drift correction. It implements three thermal-motion-based calibration methods: power spectral density, Allan variance, and Hadamard variance, with the latter providing increased robustness against drift and correlated measurement noise. The graphical user interface combines automated analysis with user-guided quality control, including reference bead selection, z-offset determination, magnet-position classification, bead-wise force estimation, and tether extension analysis. Together, these steps generate a master force calibration curve. We systematically evaluated their effects on calibration performance. The largest improvement resulted from automated filtering of problematic force plateaus using time-series characterization to identify traces affected by drift, tracking artifacts, or irregular fluctuations. Filtering reduced scatter and improved force estimation while preserving the underlying force-magnet position relationship. CALISTO thus provides a standardized, reproducible framework for reliable magnetic-tweezers force calibration.

biophysics↗

bioRxiv2026

A Planar Switching Neuron Shares Logarithmic First-Spike Latency with Its Three-Dimensional Parent But Not Its Onset Mechanism

Reducing an electronic neuron from three variables to two can preserve a familiar voltage trace while changing the underlying dynamics that create it. We examine this distinction in a piecewise-linear planar reformulation of the three-dimensional neurosimilator. The planar model produces repetitive spike-like excursions without an imposed reset and admits an exact linear solution inside each of six switching regions. Direct comparison reveals only partial agreement between the two models. Although both generate large voltage excursions and an event rate that rises with applied drive, their waveforms and frequency-current gains differ substantially. A reanalysis of the three-dimensional, near-threshold data also corrects an earlier interpretation: over the resolved interval, the first-spike latency depends logarithmically on the normalized distance from threshold, matching the behavior of the planar model. However, this common functional form does not imply a shared onset mechanism. The three-dimensional equilibrium branch reaches a saddle-node limit point, whereas the planar equilibrium remains a stable node until it exits its admissible region at a switching boundary. While a stagewise analytical reconstruction captures individual voltage excursions, crossing-time errors accumulate unless Newton corrections are applied. The reduced system is therefore a useful electronic neuron model with transparent switching dynamics, rather than a dynamically equivalent copy of its three-dimensional parent. \

biophysics↗