Geometric deep learning-guided Suzuki reaction conditions assessment for applications in medicinal chemistry.

Suzuki cross-coupling reactions are considered a valuable tool for constructing carbon-carbon bonds in small molecule drug discovery. However, the synthesis of chemical matter often represents a time-consuming and labour-intensive bottleneck. We demonstrate how machine learning methods trained on high-throughput experimentation (HTE) data can be leveraged to enable fast reaction condition selection for novel coupling partners. We show that the trained models support chemists in determining suitable catalyst-solvent-base combinations for individual transformations including an evaluation of the need for HTE screening. We introduce an algorithm for designing 96-well plates optimized towards reaction yields and discuss the model performance of zero- and few-shot machine learning. The best-performing machine learning model achieved a three-category classification accuracy of 76.3% (±0.2%) and an F 1-score for a binary classification of 79.1% (±0.9%). Validation on eight reactions revealed a receiver operating characteristic (ROC) curve (AUC) value of 0.82 (±0.07) for few-shot machine learning. On the other hand, zero-shot machine learning models achieved a mean ROC-AUC value of 0.63 (±0.16). This study positively advocates the application of few-shot machine learning-guided reaction condition selection for HTE campaigns in medicinal chemistry and highlights practical applications as well as challenges associated with zero-shot machine learning.

Comparative Analysis of Chemical Descriptors by Machine Learning Reveals Atomistic Insights into Solute-Lipid Interactions.

This study explores the research area of drug solubility in lipid excipients, an area persistently complex despite recent advancements in understanding and predicting solubility based on molecular structure. To this end, this research investigated novel descriptor sets, employing machine learning techniques to understand the determinants governing interactions between solutes and medium-chain triglycerides (MCTs). Quantitative structure-property relationships (QSPR) were constructed on an extended solubility data set comprising 182 experimental values of structurally diverse drug molecules, including both development and marketed drugs to extract meaningful property relationships. Four classes of molecular descriptors, ranging from traditional representations to complex geometrical descriptions, were assessed and compared in terms of their predictive accuracy and interpretability. These include two-dimensional (2D) and three-dimensional (3D) descriptors, Abraham solvation parameters, extended connectivity fingerprints (ECFPs), and the smooth overlap of atomic position (SOAP) descriptor. Through testing three distinct regularized regression algorithms alongside various preprocessing schemes, the SOAP descriptor enabled the construction of a superior performing model in terms of interpretability and accuracy. Its atom-centered characteristics allowed contributions to be estimated at the atomic level, thereby enabling the ranking of prevalent molecular motifs and their influence on drug solubility in MCTs. The performance on a separate test set demonstrated high predictive accuracy (RMSE = 0.50) for 2D and 3D, SOAP, and Abraham Solvation descriptors. The model trained on ECFP4 descriptors resulted in inferior predictive accuracy. Lastly, uncertainty estimations for each model were introduced to assess their applicability domains and provide information on where the models may extrapolate in chemical space and, thus, where more data may be necessary to refine a data-driven approach to predict solubility in MCTs. Overall, the presented approaches further enable computationally informed formulation development by introducing a novel in silico approach for rational drug development and prediction of dose loading in lipids.

A Bayesian approach to NMR crystal structure determination.

Nuclear Magnetic Resonance (NMR) spectroscopy is particularly well suited to determine the structure of molecules and materials in powdered form. Structure determination usually proceeds by finding the best match between experimentally observed NMR chemical shifts and those of candidate structures. Chemical shifts for the candidate configurations have traditionally been computed by electronic-structure methods, and more recently predicted by machine learning. However, the reliability of the determination depends on the errors in the predicted shifts. Here we propose a Bayesian framework for determining the confidence in the identification of the experimental crystal structure, based on knowledge of the typical errors in the electronic structure methods. We demonstrate the approach on the determination of the structures of six organic molecular crystals. We critically assess the reliability of the structure determinations, facilitated by the introduction of a visualization of the similarity between candidate configurations in terms of their chemical shifts and their structures. We also show that the commonly used values for the errors in calculated 13C shifts are underestimated, and that more accurate, self-consistently determined uncertainties make it possible to use 13C shifts to improve the accuracy of structure determinations. Finally, we extend the recently-developed ShiftML model to render it more efficient, accurate, and, most importantly, to evaluate the uncertainties in its predictions. By quantifying the confidence in structure determinations based on ShiftML predictions we further substantiate that it provides a valid replacement for first-principles calculations in NMR crystallography.

Cost-effectiveness in transient hypocalcemia post-thyroidectomy.

BACKGROUND: Three different strategies to manage transient hypocalcemia after total thyroidectomy were compared to evaluate cost-effectiveness. The reliability of total serum calcium (TSCa), ionized calcium (ICa), and intact parathyroid hormone (iPTH) were investigated to achieve this goal. METHODS: A multicenter, prospective randomized study was carried out with 169 patients. The strategies were "preventive" (oral calcium + vitamin D supplementation), "reactive" (therapy in hypocalcemia), and "predictive" (therapy if iPTH <10 pg/mL). RESULTS: TSCa had higher accuracy in identifying patients who developed hypocalcemia-related symptoms than ICa (84.6% vs 50.0%). TSCa 24 h after surgery showed 24.8% of patients with hypocalcemia, whereas TSCa 48 h after surgery identified a further 10.6% with hypocalcemia (only in the "reactive" and "predictive" groups). iPTH showed low sensitivity as a predictor of hypocalcemia. Between the 3 groups, there was no significant difference in hospitalization time or number of symptomatic hypocalcemic patients. Interestingly, the cost-per-patient was significantly different among the groups. CONCLUSIONS: None of the discussed strategies allowed for early discharge of patients without any risk of transient hypocalcemia. The "preventive" strategy was the most cost-effective, despite overtreatment.

Lateral cervical approach for open laryngeal surgery: Technical notes.

BACKGROUND: Transoral minimally invasive techniques for laryngeal cancer have been proposed to preserve healthy tissues. The aim of this study was to describe a minimally invasive procedure for all laryngectomies with/without neck dissection using a lateral cervical approach. METHODS: A monolateral or bilateral neck incision at the level of the anterior border of the sterno-cleido-mastoid muscle is performed in accordance with the side of the neck dissection. The harvesting of an anterior myocutaneous (AMC) cervical flap, including skin, platysma, fascia superficialis, anterior jugular veins, homohyoid and sternohyoid muscles is performed. A neck dissection is executed through the same skin incision. Contraindication is represented by extralaryngeal spread with the involvement of the strap muscles. Supra-hyoid, sternohyoid, and homohyoid muscles and hyoid bone are preserved. CONCLUSION: Open partial/total laryngectomies through a lateral cervical approach are feasible and safe. This surgical access with an AMS cervical flap allows to simultaneously perform neck dissections and the removal of the entire laryngeal specimen.

Automatic selection of atomic fingerprints and reference configurations for machine-learning potentials.

Machine learning of atomic-scale properties is revolutionizing molecular modeling, making it possible to evaluate inter-atomic potentials with first-principles accuracy, at a fraction of the costs. The accuracy, speed, and reliability of machine learning potentials, however, depend strongly on the way atomic configurations are represented, i.e., the choice of descriptors used as input for the machine learning method. The raw Cartesian coordinates are typically transformed in "fingerprints," or "symmetry functions," that are designed to encode, in addition to the structure, important properties of the potential energy surface like its invariances with respect to rotation, translation, and permutation of like atoms. Here we discuss automatic protocols to select a number of fingerprints out of a large pool of candidates, based on the correlations that are intrinsic to the training data. This procedure can greatly simplify the construction of neural network potentials that strike the best balance between accuracy and computational efficiency and has the potential to accelerate by orders of magnitude the evaluation of Gaussian approximation potentials based on the smooth overlap of atomic positions kernel. We present applications to the construction of neural network potentials for water and for an Al-Mg-Si alloy and to the prediction of the formation energies of small organic molecules using Gaussian process regression.

Mapping uncharted territory in ice from zeolite networks to ice structures.

Ice is one of the most extensively studied condensed matter systems. Yet, both experimentally and theoretically several new phases have been discovered over the last years. Here we report a large-scale density-functional-theory study of the configuration space of water ice. We geometry optimise 74,963 ice structures, which are selected and constructed from over five million tetrahedral networks listed in the databases of Treacy, Deem, and the International Zeolite Association. All prior knowledge of ice is set aside and we introduce "generalised convex hulls" to identify configurations stabilised by appropriate thermodynamic constraints. We thereby rediscover all known phases (I-XVII, i, 0 and the quartz phase) except the metastable ice IV. Crucially, we also find promising candidates for ices XVIII through LI. Using the "sketch-map" dimensionality-reduction algorithm we construct an a priori, navigable map of configuration space, which reproduces similarity relations between structures and highlights the novel candidates. By relating the known phases to the tractably small, yet structurally diverse set of synthesisable candidate structures, we provide an excellent starting point for identifying formation pathways.

Complications of orbital floor repair with silastic sheet: the skin fistula.

Treatment of orbital floor fracture is a subject of great interest in maxillofacial surgery. Many materials have been described for its reconstruction.In this article, the authors report a case of a patient who, 7 years from a previous orbital floor fracture and treatment with silastic sheet, presented herself to their clinic for the failure of the material used for its reconstruction and a skin fistula.Orbital floor repair with silastic sheet is an old method that no one uses anymore, but we still observe cases of late complications with this material. So a fine knowledge of silastic sheet complications is needed for young surgeons.The authors report the case and perform a literature review about the use of more modern biomaterials for orbital floor reconstruction.

Management of clinically negative neck in maxillary carcinoma.

Maxillary cancers include neoplasms arising in both maxillary sinus and oral cavity (upper alveolar ridge, hard palate) according to the American Joint Committee on Cancer. Although it is universally accepted that the combination of surgery and radiotherapy seems to be the treatment of choice, there is no accordance about the treatment of clinically negative neck. We retrospectively analyzed 20 patients with maxillary sinus cancer and 37 with an upper alveolar ridge or hard palate cancer, evaluating the incidence of N-disease and the recurrence at local site. On the basis of our findings, we can affirm that elective treatment of the neck in maxillary carcinoma is not recommended. Considering only squamous cell carcinoma, cervical node metastases are most frequent in case of tumors staged as T1 o T2. High-grade squamous cell carcinomas seem to be related to a higher incidence of nodal involvement. T recurrence has demonstrated to be the most frequent neoplastic event, so that radical surgery is considered one of the most important prognostic factors. Nevertheless, other prospective studies are necessary.

Scapula free flap for complex maxillofacial reconstruction.

INTRODUCTION: Composite tissue defects of the mandible and maxilla, after resection of head and neck malignancies, osteoradionecrosis, malformations, or traumas, cause functional and aesthetic problems. Nowadays, microvascular free flaps represent the main choice for the reconstruction of these defects. Among the various flaps proposed, the scapula flap has favorable characteristics that make it suitable for bone, soft tissue, or combined defects. MATERIALS: We report 7 cases of reconstruction of complex maxillofacial defects with subscapular system flaps. The patients treated had Romberg syndrome (1 case), malignant tumors (5 cases), and result of previous trauma (1 case).Location of deficit was the maxilla (3 cases), the mandible (2 case), the ethmoidal-maxillary region (1 case) and the upper and middle thirds of the face in the last case. METHODS: In 2 cases, a parascapular system flap was used; in 5 cases, a composite flap with latissimus dorsi muscle and scapular bone. RESULTS: Neither failure of the harvested flaps nor complications in the donor site were evidenced. A good aesthetic and functional outcome was obtained in all cases. DISCUSSION: : Many free flaps have been proposed for the reconstruction of defects in the maxillofacial region such as fibula, deep circumflex iliac artery, scapula, among the bone flaps; and forearm, rectus abdominis, and anterolateral thigh, among the soft tissue flaps. The choice of the flap to use depends on the length of the bone defect and the amount of soft tissues required. The subscapular system has the advantage of providing different flaps based on the same pedicle. The osteofasciocutaneous scapular free flap, in particular, allows wide mobility of soft tissues (parascapular flap) with respect to its bone component (scapular bone), resulting suitable for defects of large size involving both the soft tissues and the bone. CONCLUSIONS: Although the fibula flap and the deep circumflex iliac artery flap remain the first choice for bone reconstructions of the mandible and maxilla, the scapula flap has some features that make its use extremely advantageous in some circumstances. In particular, we advocate the use of the osteomuscular latissimus dorsi-scapula flap for reconstruction of large-volume defects involving the bone and soft tissues, whereas fasciocutaneous parascapular flaps represent a valid alternative to forearm flap and anterolateral thigh flap in the reconstruction of soft tissue defects.