Journal: bioRxiv

Article Title: 13 C-Metabolic Flux Analysis of Substrate Co-Metabolism in the Methylotrophic Acetogen Eubacterium limosum

doi: 10.64898/2026.10.12.758492

Figure Lengend Snippet: a) Growth plot for E. limosum grown on glucose in defined medium supplemented with cysteine. Mass Isotopologue Distributions (MID) for proteinogenic amino acids obtained from E. limosum grown with [3- 13 C] cysteine, with methanol (b) or glucose (c). Standard deviation for MID measurements were less than 1% for glucose-grown cells, and less than 6% for methanol-grown cells.

Article Snippet: Wild-type Eubacterium limosum (ATCC 8486) was used for batch experiments in Hungate tubes comparing potential sulfur sources and for tracer experiments with labeled cysteine.

Techniques: Standard Deviation

Journal: Nature Communications

Article Title: Time-lapse RESOLFT nanoscopy empowered by deep learning

doi: 10.1038/s41467-026-78580-z

Figure Lengend Snippet: a Sequence of MoNaLISA imaging. Black arrows shown between 488 nm read-out foci represent the scanning pattern used during imaging. b MoNaLISA imaging of fixed U2OS cells endogenously expressing vimentin-rsEGFP2 for: 80-90% ON-switching (left), 10-20% ON-switching (middle), and 10–20% ON-switching denoised with our proposed method. Representative images from a test dataset of n = 5 cells with similar results. Scale bar, 1 µm. c Denoised predictions of 10–20% ON-switching MoNaLISA for HDN, N2V and SN2N as unsupervised methods (top row) and HDN sup , UNet-RCAN and CARE as supervised ones (bottom row). Scale bar, 1 µm. d Resolution analysis of denoising results for unsupervised (top row) and supervised methods (bottom row). Line profile example with normalized intensity (left column), single-image Fourier Ring Correlation (mean ± std, N = 3 images, middle column), and FWHM fit of line profiles of 30 filaments. Box plots show the mean (centre line), 25th and 75th percentiles (box bounds), and whiskers show data points within 1.5 times the interquartile range. e Live HeLa cells expressing fusion protein HIV Gag-IM-rsEGFP. From left to right: enhanced confocal (no OFF-switching), RESOLFT with 90% ON-switching, RESOLFT with 10% ON-switching and predictions of the 10% ON-switching RESOLFT image with unsupervised methods: SN2N, N2V and HDN. Scale bar, 1 µm f Single-image Fourier Ring Correlation (mean ± std, N = 8 images). Color bar indicates normalized image intensity independently scaled for display.

Article Snippet: Wild-type HeLa (ATCC CCL-2) and U2OS (ATCC HTB-96) cells and U2OS with endogenously tagged vimentin-rsEGFP2 (ref. Monalisa) were cultured in DMEM (Thermo Fisher Scientific; 41966029) supplemented with 10% (vol/vol) FBS (Thermo Fisher Scientific; 10270106) and 1% penicillin–streptomycin (Sigma Aldrich; P4333) and maintained at 37 °C and 5% CO2 in a humidified incubator.

Techniques: Sequencing, Imaging, Expressing, Analysis, Confocal

Journal: Nature Communications

Article Title: Time-lapse RESOLFT nanoscopy empowered by deep learning

doi: 10.1038/s41467-026-78580-z

Figure Lengend Snippet: a Temporal resolution dependency over the scanning pixel size. Frame rate calculated for a multi-foci period of 625 nm and a dwell time of 5 ms. b Training setup for Fast-MoNaLISA. Starting from a Nyquist-sampled image (the GT), we select a subset of pixels to create a subsampled input of a given sampling ratio S (75%, 50% or 25%). Images are representative examples from the test dataset ( n = 3 cells). Scale bar, 1 µm. c Fast-MoNaLISA results examples on U2OS cells endogenously expressing vimentin-rsEGFP2, across sampling ratios. The line profile across the orange arrows is reported in orange. Scale bar, 1 µm. d Relative FRC between predictions and GT on the vimentin dataset (mean ± std, n = 3 cells). e PSNR and SSIM performances for all sampling ratios on the vimentin dataset ( n = 3 cells). Centre line: mean, box bounds: 25th and 75th percentile, whiskers: extreme data points within 1.5 time the interquartile range. f Diffraction-limited confocal image, and fast-MoNaLISA results examples on live HeLa cells expressing fusion protein HIV Gag-IM-rsEGFP. White arrows highlight an improvement in the 50% sampling prediction over the 25% sampling prediction. Similar results are found on the whole test dataset ( n = 17 cells). Scale bar, 1 µm g Single-image Fourier Ring Correlation on the Gag-IM particles dataset (mean ± std, n = 17 images). Color bar indicates normalized image intensity independently scaled for display.

Article Snippet: Wild-type HeLa (ATCC CCL-2) and U2OS (ATCC HTB-96) cells and U2OS with endogenously tagged vimentin-rsEGFP2 (ref. Monalisa) were cultured in DMEM (Thermo Fisher Scientific; 41966029) supplemented with 10% (vol/vol) FBS (Thermo Fisher Scientific; 10270106) and 1% penicillin–streptomycin (Sigma Aldrich; P4333) and maintained at 37 °C and 5% CO2 in a humidified incubator.

Techniques: Sampling, Expressing, Confocal