Trustworthy Data and AI Environments for Clinical Prediction: Application to Crisis-Risk in People With Depression.

Depression is a common mental health condition that often occurs in association with other chronic illnesses, and varies considerably in severity. Electronic Health Records (EHRs) contain rich information about a patient's medical history and can be used to train, test and maintain predictive models to support and improve patient care. This work evaluated the feasibility of implementing an environment for predicting mental health crisis among people living with depression based on both structured and unstructured EHRs. A large EHR from a mental health provider, Mersey Care, was pseudonymised and ingested into the Natural Language Processing (NLP) platform CogStack, allowing text content in binary clinical notes to be extracted. All unstructured clinical notes and summaries were semantically annotated by MedCAT and BioYODIE NLP services. Cases of crisis in patients with depression were then identified. Random forest models, gradient boosting trees, and Long Short-Term Memory (LSTM) networks, with varying feature arrangement, were trained to predict the occurrence of crisis. The results showed that all the prediction models can use a combination of structured and unstructured EHR information to predict crisis in patients with depression with good and useful accuracy. The LSTM network that was trained on a modified dataset with only 1000 most-important features from the random forest model with temporality showed the best performance with a mean AUC of 0.901 and a standard deviation of 0.006 using a training dataset and a mean AUC of 0.810 and 0.01 using a hold-out test dataset. Comparing the results from the technical evaluation with the views of psychiatrists shows that there are now opportunities to refine and integrate such prediction models into pragmatic point-of-care clinical decision support tools for supporting mental healthcare delivery.

Untangling the threads of cellulose mercerization.

Naturally occurring plant cellulose, our most abundant renewable resource, consists of fibers of long polymer chains that are tightly packed in parallel arrays in either of two crystal phases collectively referred to as cellulose I. During mercerization, a process that involves treatment with sodium hydroxide, cellulose goes through a conversion to another crystal form called cellulose II, within which every other chain has remarkably changed direction. We designed a neutron diffraction experiment with deuterium labelling in order to understand how this change of cellulose chain direction is possible. Here we show that during mercerization of bacterial cellulose, chains fold back on themselves in a zigzag pattern to form crystalline anti-parallel domains. This result provides a molecular level understanding of one of the most widely used industrial processes for improving cellulosic materials.

Capturing the Catalytic Proton of Dihydrofolate Reductase: Implications for General Acid-Base Catalysis.

Acid-base catalysis, which involves one or more proton transfer reactions, is a chemical mechanism commonly employed by many enzymes. The molecular basis for catalysis is often derived from structures determined at the optimal pH for enzyme activity. However, direct observation of protons from experimental structures is quite difficult; thus, a complete mechanistic description for most enzymes remains lacking. Dihydrofolate reductase (DHFR) exemplifies general acid-base catalysis, requiring hydride transfer and protonation of its substrate, DHF, to form the product, tetrahydrofolate (THF). Previous X-ray and neutron crystal structures coupled with theoretical calculations have proposed that solvent mediates the protonation step. However, visualization of a proton transfer has been elusive. Based on a 2.1 Å resolution neutron structure of a pseudo-Michaelis complex of E. coli DHFR determined at acidic pH, we report the direct observation of the catalytic proton and its parent solvent molecule. Comparison of X-ray and neutron structures elucidated at acidic and neutral pH reveals dampened dynamics at acidic pH, even for the regulatory Met20 loop. Guided by the structures and calculations, we propose a mechanism where dynamics are crucial for solvent entry and protonation of substrate. This mechanism invokes the release of a sole proton from a hydronium (H3O+) ion, its pathway through a narrow channel that sterically hinders the passage of water, and the ultimate protonation of DHF at the N5 atom.

Structural plasticity of SARS-CoV-2 3CL Mpro active site cavity revealed by room temperature X-ray crystallography.

The COVID-19 disease caused by the SARS-CoV-2 coronavirus has become a pandemic health crisis. An attractive target for antiviral inhibitors is the main protease 3CL Mpro due to its essential role in processing the polyproteins translated from viral RNA. Here we report the room temperature X-ray structure of unliganded SARS-CoV-2 3CL Mpro, revealing the ligand-free structure of the active site and the conformation of the catalytic site cavity at near-physiological temperature. Comparison with previously reported low-temperature ligand-free and inhibitor-bound structures suggest that the room temperature structure may provide more relevant information at physiological temperatures for aiding in molecular docking studies.

BraggNet: integrating Bragg peaks using neural networks.

Neutron crystallography offers enormous potential to complement structures from X-ray crystallography by clarifying the positions of low-Z elements, namely hydrogen. Macromolecular neutron crystallography, however, remains limited, in part owing to the challenge of integrating peak shapes from pulsed-source experiments. To advance existing software, this article demonstrates the use of machine learning to refine peak locations, predict peak shapes and yield more accurate integrated intensities when applied to whole data sets from a protein crystal. The artificial neural network, based on the U-Net architecture commonly used for image segmentation, is trained using about 100 000 simulated training peaks derived from strong peaks. After 100 training epochs (a round of training over the whole data set broken into smaller batches), training converges and achieves a Dice coefficient of around 65%, in contrast to just 15% for negative control data sets. Integrating whole peak sets using the neural network yields improved intensity statistics compared with other integration methods, including k-nearest neighbours. These results demonstrate, for the first time, that neural networks can learn peak shapes and be used to integrate Bragg peaks. It is expected that integration using neural networks can be further developed to increase the quality of neutron, electron and X-ray crystallography data.

Improving the accuracy and resolution of neutron crystallographic data by three-dimensional profile fitting of Bragg peaks in reciprocal space.

Neutron crystallography is a powerful technique for directly visualizing the locations of H atoms in biological macromolecules. This information has provided key new insights into enzyme mechanisms, ligand binding and hydration. However, despite the importance of this information, the application of neutron crystallography in biology has been limited by the relatively low flux of available neutron beams and the large incoherent neutron scattering from hydrogen, both of which contribute to weak diffraction data with relatively low signal-to-background ratios. A method has been developed to fit weak data based on three-dimensional profile fitting of Bragg peaks in reciprocal space by an Ikeda-Carpenter function with a bivariate Gaussian. When applied to data collected from three different proteins, three-dimensional profile fitting yields intensities with higher correlation coefficients (CC1/2) at high resolutions, decreased Rfree factors, extended resolutions and improved nuclear density maps. Importantly, additional features are revealed in nuclear density maps that may provide additional scientific information. These results suggest that three-dimensional profile fitting will help to extend the capabilities of neutron macromolecular crystallography.

Evaluation of models determined by neutron diffraction and proposed improvements to their validation and deposition.

The Protein Data Bank (PDB) contains a growing number of models that have been determined using neutron diffraction or a hybrid method that combines X-ray and neutron diffraction. The advantage of neutron diffraction experiments is that the positions of all atoms can be determined, including H atoms, which are hardly detectable by X-ray diffraction. This allows the determination of protonation states and the assignment of H atoms to water molecules. Because neutrons are scattered differently by hydrogen and its isotope deuterium, neutron diffraction in combination with H/D exchange can provide information on accessibility, dynamics and chemical lability. In this study, the deposited data, models and model-to-data fit for all PDB entries that used neutron diffraction as the source of experimental data have been analysed. In many cases, the reported Rwork and Rfree values were not reproducible. In such cases, the model and data files were analysed to identify the reasons for this mismatch. The issues responsible for the discrepancies are summarized and explained. The analysis unveiled limitations to the annotation, deposition and validation of models and data, and a lack of community-wide accepted standards for the description of neutron models and data, as well as deficiencies in current model refinement tools. Most of the issues identified concern the handling of H atoms. Since the primary use of neutron macromolecular crystallography is to locate and directly visualize H atoms, it is important to address these issues, so that the deposited neutron models allow the retrieval of the maximum amount of information with the smallest effort of manual intervention. A path forward to improving the annotation, validation and deposition of neutron models and hybrid X-ray and neutron models is suggested.

Tension wood structure and morphology conducive for better enzymatic digestion.

BACKGROUND: Tension wood is a type of reaction wood in response to bending or leaning stem as a corrective growth process. Tension wood is formed by both natural and man-made processes. Most attractively, tension wood contains higher glucan content and undergoes higher enzymatic conversion to fermentable sugars. Here, we have employed structural techniques, small-angle neutron scattering (SANS) and wide-angle X-ray diffraction (WAXD) to elucidate structural and morphological aspects of tension wood conducive to higher sugar yields. RESULTS: Small-angle neutron scattering data exhibited a tri-modal distribution of the fibril cross-sectional dimension. The smallest size, 22 Å observed in all samples concurred with the WAXD results of the control and opposite side samples. This smallest and the most abundant occurring size was interpreted as the cellulose elementary microfibril diameter. The intermediate size of 45 Å, which is most pronounced in the tension side sample and consistent with WAXD results for tension side sample, indicates association of neighboring elementary microfibrils to form larger crystallite bundles. The largest size 61 Å observed by SANS was however not observed by WAXD and therefore associated to mesopores. CONCLUSIONS: Structure and morphology of tension wood is different from control wood. Cellulose crystallinity increases, lignin content is lower and the appearance of mesopores with 61 Å diameter is observed. Despite the presence of higher crystalline cellulose content in tension side, the lower lignin content and may be combined with the abundance of mesopores, substantially improves enzyme accessibility leading to higher yields in cellulose digestion.

Neutron scattering in the biological sciences: progress and prospects.

The scattering of neutrons can be used to provide information on the structure and dynamics of biological systems on multiple length and time scales. Pursuant to a National Science Foundation-funded workshop in February 2018, recent developments in this field are reviewed here, as well as future prospects that can be expected given recent advances in sources, instrumentation and computational power and methods. Crystallography, solution scattering, dynamics, membranes, labeling and imaging are examined. For the extraction of maximum information, the incorporation of judicious specific deuterium labeling, the integration of several types of experiment, and interpretation using high-performance computer simulation models are often found to be particularly powerful.