A parametric approach for molecular encodings using multilevel atomic neighborhoods applied to peptide classification.

Exploring new ways to represent and discover organic molecules is critical to the development of new therapies. Fingerprinting algorithms are used to encode or machine-read organic molecules. Molecular encodings facilitate the computation of distance and similarity measurements to support tasks such as similarity search or virtual screening. Motivated by the ubiquity of carbon and the emerging structured patterns, we propose a parametric approach for molecular encodings using carbon-based multilevel atomic neighborhoods. It implements a walk along the carbon chain of a molecule to compute different representations of the neighborhoods in the form of a binary or numerical array that can later be exported into an image. Applied to the task of binary peptide classification, the evaluation was performed by using forty-nine encodings of twenty-nine data sets from various biomedical fields, resulting in well over 1421 machine learning models. By design, the parametric approach is domain- and task-agnostic and scopes all organic molecules including unnatural and exotic amino acids as well as cyclic peptides. Applied to peptide classification, our results point to a number of promising applications and extensions. The parametric approach was developed as a Python package (cmangoes), the source code and documentation of which can be found at https://github.com/ghattab/cmangoes and https://doi.org/10.5281/zenodo.7483771.

A new technological approach in diagnostic pathology: mass spectrometry imaging-based metabolomics for biomarker detection in urachal cancer.

Urachal adenocarcinomas (UrC) are rare but aggressive. Despite being of profound therapeutic relevance, UrC cannot be differentiated by histomorphology alone from other adenocarcinomas of differential diagnostic importance. As no reliable tissue-based diagnostic biomarkers are available, we aimed to detect such by integrating mass-spectrometry imaging-based metabolomics and digital pathology, thus allowing for a multimodal approach on the basis of spatial information. To achieve this, a cohort of UrC (n = 19) and colorectal adenocarcinomas (CRC, n = 27) as the differential diagnosis of highest therapeutic relevance was created, tissue micro-arrays (TMAs) were constructed, and pathological data was recorded. Hematoxylin and eosin (H&E) stained tissue sections were scanned and annotated, enabling an automized discrimination of tumor and non-tumor areas after training of an adequate algorithm. Spectral information within tumor regions, obtained via matrix-assisted laser desorption/ionization (MALDI)-Orbitrap-mass spectrometry imaging (MSI), were subsequently extracted in an automated workflow. On this basis, metabolic differences between UrC and CRC were revealed using machine learning algorithms. As a result, the study demonstrated the feasibility of MALDI-MSI for the evaluation of FFPE tissue in UrC and CRC with the potential to combine spatial metabolomics data with annotated histopathological data from digitalized H&E slides. The detected Area under the curve (AUC) of 0.94 in general and 0.77 for the analyte taurine alone (diagnostic accuracy for taurine: 74%) makes the technology a promising tool in this differential diagnostic dilemma situation. Although the data has to be considered as a proof-of-concept study, it presents a new adoption of this technology that has not been used in this scenario in which reliable diagnostic biomarkers (such as immunohistochemical markers) are currently not available.

Routine Endovascular Treatment With a Stent Graft for Access-Site and Access-Related Vascular Injury in Transfemoral Transcatheter Aortic Valve Implantation.

BACKGROUND: Access-site and access-related vascular injury (ASARVI) is still a major limiting factor in transcatheter aortic valve implantation and affects the outcome of patients. Management strategies for ASARVI include manual compression, stent grafts, and vascular surgery. We hypothesized that the standard use of a self-expanding stent graft for the management of ASARVI is feasible and safe. METHODS AND RESULTS: Of 407 patients treated by transfemoral transcatheter aortic valve implantation, 110 experienced ASARVI (27.0%). Of these, 96 (87.3%) were managed by the implantation of a self-expanding nitinol stent graft. In the majority of patients, minor vascular complications triggered the implantation of a stent graft (86.5%), mainly because of bleeding (90.6%) and dissection (5.2%) of the common femoral artery with high rates of primary treatment success (97.9%). Patients receiving stent grafts were more often female (62.2 versus 45.6%, P<0.01), had higher body mass indices (27.8±6.7 versus 25.7±4.7, P=0.01), and suffered more often from diabetes mellitus (34.4 versus 24.5%, P=0.04). Angiographic assessment after a median follow-up of 345 days (interquartile range, 23-745 days) revealed only one patient with moderate, asymptomatic instent-stenosis (1.0%). Compared with a propensity score-matched cohort of patients without ASARVI, stented patients had comparable long-term mortality, despite the occurrence of a vascular complication (1-year mortality: 17.7% versus 26.6%; stent versus matched cohort, respectively; P=0.1). CONCLUSIONS: Routine use of a self-expanding nitinol stent graft in selected patients experiencing ASARVI after transcatheter aortic valve implantation is feasible, safe, and associated with favorable short- and midterm clinical outcome.