Explore the vast range of titles available.

The size and shape of peptide ions in the gas phase are an under-explored dimension for mass spectrometry-based proteomics. To investigate the nature and utility of the peptide collisional cross section (CCS) space, we measure more than a million data points from whole-proteome digests of five organisms with trapped ion mobility spectrometry (TIMS) and parallel accumulation-serial fragmentation (PASEF). The scale and precision (CV < 1%) of our data is sufficient to train a deep recurrent neural network that accurately predicts CCS values solely based on the peptide sequence. Cross section predictions for the synthetic ProteomeTools peptides validate the model within a 1.4% median relative error ( R  > 0.99). Hydrophobicity, proportion of prolines and position of histidines are main determinants of the cross sections in addition to sequence-specific interactions. CCS values can now be predicted for any peptide and organism, forming a basis for advanced proteomics workflows that make full use of the additional information. Proteomics has been advanced by algorithms that can predict different peptide features, but predicting peptide collisional cross sections (CCS) has remained challenging. Here, the authors measure over one million CCS values of tryptic peptides and develop a deep learning model for peptide CCS prediction.

We show that deep convolutional neural networks combined with nonlinear dimension reduction enable reconstructing biological processes based on raw image data. We demonstrate this by reconstructing the cell cycle of Jurkat cells and disease progression in diabetic retinopathy. In further analysis of Jurkat cells, we detect and separate a subpopulation of dead cells in an unsupervised manner and, in classifying discrete cell cycle stages, we reach a sixfold reduction in error rate compared to a recent approach based on boosting on image features. In contrast to previous methods, deep learning based predictions are fast enough for on-the-fly analysis in an imaging flow cytometer. The interpretation of information-rich, high-throughput single-cell data is a challenge requiring sophisticated computational tools. Here the authors demonstrate a deep convolutional neural network that can classify cell cycle status on-the-fly.

Proteins carry out the vast majority of functions in all biological domains, but for technological reasons their large-scale investigation has lagged behind the study of genomes. Since the first essentially complete eukaryotic proteome was reported 1 , advances in mass-spectrometry-based proteomics 2 have enabled increasingly comprehensive identification and quantification of the human proteome 3 – 6 . However, there have been few comparisons across species 7 , 8 , in stark contrast with genomics initiatives 9 . Here we use an advanced proteomics workflow—in which the peptide separation step is performed by a microstructured and extremely reproducible chromatographic system—for the in-depth study of 100 taxonomically diverse organisms. With two million peptide and 340,000 stringent protein identifications obtained in a standardized manner, we double the number of proteins with solid experimental evidence known to the scientific community. The data also provide a large-scale case study for sequence-based machine learning, as we demonstrate by experimentally confirming the predicted properties of peptides from Bacteroides uniformis . Our results offer a comparative view of the functional organization of organisms across the entire evolutionary range. A remarkably high fraction of the total proteome mass in all kingdoms is dedicated to protein homeostasis and folding, highlighting the biological challenge of maintaining protein structure in all branches of life. Likewise, a universally high fraction is involved in supplying energy resources, although these pathways range from photosynthesis through iron sulfur metabolism to carbohydrate metabolism. Generally, however, proteins and proteomes are remarkably diverse between organisms, and they can readily be explored and functionally compared at www.proteomesoflife.org . An advanced proteomics workflow is used to identify 340,000 proteins from 100 taxonomically diverse species, providing a comparative view of proteomes across the evolutionary range.

Because of their resilience, Time-of-Flight (ToF) cameras are now essential components in scientific and industrial settings. This paper outlines the essential factors for modeling 3D ToF cameras, with specific emphasis on analyzing the phenomenon known as “wiggling”. Through our investigation, we demonstrate that wiggling not only causes systematic errors in distance measurements, but also introduces periodic fluctuations in statistical measurement uncertainty, which compounds the dependence on the signal-to-noise ratio (SNR). Armed with this knowledge, we developed a new 3D camera model, which we then made computationally tractable. To illustrate and evaluate the model, we compared measurement data with simulated data of the same scene. This allowed us to individually demonstrate various effects on the signal-to-noise ratio, reflectivity, and distance.

Access to large, annotated samples represents a considerable challenge for training accurate deep-learning models in medical imaging. Although at present transfer learning from pre-trained models can help with cases lacking data, this limits design choices and generally results in the use of unnecessarily large models. Here we propose a self-supervised training scheme for obtaining high-quality, pre-trained networks from unlabelled, cross-modal medical imaging data, which will allow the creation of accurate and efficient models. We demonstrate the utility of the scheme by accurately predicting retinal thickness measurements based on optical coherence tomography from simple infrared fundus images. Subsequently, learned representations outperformed advanced classifiers on a separate diabetic retinopathy classification task in a scenario of scarce training data. Our cross-modal, three-stage scheme effectively replaced 26,343 diabetic retinopathy annotations with 1,009 semantic segmentations on optical coherence tomography and reached the same classification accuracy using only 25% of fundus images, without any drawbacks, since optical coherence tomography is not required for predictions. We expect this concept to apply to other multimodal clinical imaging, health records and genomics data, and to corresponding sample-starved learning problems. The thickness of the retina is an important medical indicator for diabetic retinopathy. Holmberg and colleagues present a self-supervised deep-learning method that uses cross-modal data to predict retinal thickness maps from easily obtainable fundus images.

To the Editor: Conradie et al. (Sept. 1 issue) 1 report impressive results of the ZeNix trial of a bedaquiline–pretomanid–linezolid regimen for drug-resistant tuberculosis, with doses of linezolid that were lower than the 1200-mg dose that has been associated with adverse events. In this trial performed in South Africa, the country of Georgia, Moldova, and Russia, 9 of the 181 participants (5.0%) had bedaquiline resistance (minimum inhibitory concentration, 2 to 4 μg per milliliter) at baseline. Despite reported 99.8% adherence to the trial regimen, 6 of 9 participants (67%) with bedaquiline-resistant tuberculosis at baseline had a favorable outcome, as compared with . . .

Tuberculosis is a bacterial infectious disease that is mainly transmitted from human to human via infectious aerosols. Currently, tuberculosis is the leading cause of death by an infectious disease world-wide. In the past decade, the number of patients affected by tuberculosis has increased by ~20 percent and the emergence of drug-resistant strains of challenges the goal of elimination of tuberculosis in the near future. For the last 50 years, management of patients with tuberculosis has followed a standardized management approach. This standardization neglects the variation in human susceptibility to infection, immune response, the pharmacokinetics of drugs, and the individual duration of treatment needed to achieve relapse-free cure. Here we propose a package of precision medicine-guided therapies that has the prospect to drive clinical management decisions, based on both host immunity and strains genetics. Recently, important scientific discoveries and technological advances have been achieved that provide a perspective for individualized rather than standardized management of patients with tuberculosis. For the individual selection of best medicines and host-directed therapies, personalized drug dosing, and treatment durations, physicians treating patients with tuberculosis will be able to rely on these advances in systems biology and to apply them at the bedside.

Background: Pediatric respiratory disorders frequently necessitate clinical evaluation, often during sleep. Traditional polysomnography (PSG), while the gold standard for sleep-related respiratory assessment, is resource-intensive and can cause discomfort, particularly in children. Therefore, in a prior published study, we designed and technically validated a video-based prototype for contactless monitoring of respiratory movements. Objective: Our present study aimed to clinically validate the contactless monitoring prototype in pediatric patients, with a primary focus on detecting respiratory rate and identifying abnormal breathing patterns. Methods: Twenty-seven pediatric patients (aged 6 months to 12 years) were recruited from a pediatric sleep laboratory. To monitor thoracoabdominal movements in real time, the prototype employed a time-of-flight camera and a 3D imaging module, coupled with artificial-intelligence-based determination of the region of interest (ROI). Respiratory rates obtained from the prototype were compared to simultaneously recorded PSG data. Data were collected under various conditions, including different sleeping positions. A total of 296 h of respiratory data were acquired, of which selected 60 s segments (54 during N3 sleep and 27 during REM sleep) were analyzed using the prototype and compared with PSG-derived respiratory parameters. Conclusion: The contactless prototype demonstrates that reliable and non-invasive respiratory monitoring is feasible in pediatric patients. It enables accurate detection of respiratory rate as well as abnormal breathing patterns under routine clinical conditions, while reducing patient burden compared with conventional approaches. Its usability and minimal patient discomfort suggest potential for broader clinical adoption. Future work should focus on full-night recordings across all sleep stages and the development of automated data analysis pipelines to facilitate routine clinical implementation.

The treatment of drug-resistant Mycobacterium tuberculosis relies on complex antibiotic therapy. Inadequate antibiotic exposure can lead to treatment failure, acquired drug resistance, and an increased risk of adverse events. Therapeutic drug monitoring (TDM) can be used to optimize the antibiotic exposure. Therefore, we aimed to develop a single-run multiplex assay using high-performance liquid chromatography–mass spectrometry (HPLC–MS) for TDM of patients with multidrug-resistant, pre-extensively drug-resistant and extensively drug-resistant tuberculosis. A target profile for sufficient performance, based on the intended clinical application, was established and the assay was developed accordingly. Antibiotics were analyzed on a zwitterionic hydrophilic interaction liquid chromatography column and a triple quadrupole mass spectrometer using stable isotope-labeled internal standards. The assay was sufficiently sensitive to monitor drug concentrations over five half-lives for rifampicin, rifabutin, levofloxacin, moxifloxacin, bedaquiline, linezolid, clofazimine, terizidone/cycloserine, ethambutol, delamanid, pyrazinamide, meropenem, prothionamide, and para-amino salicylic acid (PAS). Accuracy and precision were sufficient to support clinical decision making (≤±15% in clinical samples and ±20–25% in spiked samples, with 80% of future measured concentrations predicted to fall within ±40% of nominal concentrations). The method was applied in the TDM of two patients with complex drug-resistant tuberculosis. All relevant antibiotics from their regimens could be quantified and high-dose therapy was initiated, followed by microbiological conversion. In conclusion, we developed a multiplex assay that enables TDM of the relevant first- and second-line anti-tuberculosis medicines in a single run and was able to show its applicability in TDM of two drug-resistant tuberculosis patients.