Deep Aramaic: Towards a synthetic data paradigm enabling machine learning in epigraphy.

Epigraphy is witnessing a growing integration of artificial intelligence, notably through its subfield of machine learning (ML), especially in tasks like extracting insights from ancient inscriptions. However, scarce labeled data for training ML algorithms severely limits current techniques, especially for ancient scripts like Old Aramaic. Our research pioneers an innovative methodology for generating synthetic training data tailored to Old Aramaic letters. Our pipeline synthesizes photo-realistic Aramaic letter datasets, incorporating textural features, lighting, damage, and augmentations to mimic real-world inscription diversity. Despite minimal real examples, we engineer a dataset of 250 000 training and 25 000 validation images covering the 22 letter classes in the Aramaic alphabet. This comprehensive corpus provides a robust volume of data for training a residual neural network (ResNet) to classify highly degraded Aramaic letters. The ResNet model demonstrates 95% accuracy in classifying real images from the 8th century BCE Hadad statue inscription. Additional experiments validate performance on varying materials and styles, proving effective generalization. Our results validate the model's capabilities in handling diverse real-world scenarios, proving the viability of our synthetic data approach and avoiding the dependence on scarce training data that has constrained epigraphic analysis. Our innovative framework elevates interpretation accuracy on damaged inscriptions, thus enhancing knowledge extraction from these historical resources.

Effect of the isotiazole adjuvants in combination with cisplatin in chemotherapy of neuroepithelial tumors: experimental results and modeling.

Chemotherapy is one of the main treatment options for cancer, but it is usually accompanied with negative side effects. The classical drugs combination with synergistic adjuvants can be the solution to this problem, allowing reducing therapeutic dose. Elucidating the mechanism of adjuvant action is of key importance for the selection of the optimal agent. Here we examine the system drug-adjuvant to explain the observed effect in practice. We used the first line drug cisplatin. Morpholinium and 4-methylpiperazinium 4,5-dichloro isothiazol-3-carboxylates were selected as adjuvants. The study of the cisplatin-adjuvant system was carried out by quantum chemical modeling using DFT. It turned out that adjuvants form conjugates with cisplatin that lead to the relocation of frontier molecular orbitals as well as increase of conjugate's dipole moment. It resulted in change of the interaction character with DNA and increase of the bioactivity of the system. The data obtained are the basis for expanding the studies to include other drugs and adjuvants. Oncologists will have opportunity to use "classical" chemotherapy drugs in combination with synergists for those patients who have not been previously recommended to such a treatment because of pronounced toxic side effects.

A Current Loop Model for the Fast Simulation of Ferrofluids.

Ferrofluids are oil-based liquids containing magnetic particles that interact with magnetic fields without solidifying. Leveraging the exploration of new applications of these promising materials (such as in optics, medicine and engineering) requires high fidelity modeling and simulation capabilities in order to accurately explore ferrofluids in silico. While recent work addressed the macroscopic simulation of large-scale ferrofluids using smoothed-particle hydrodynamics (SPH), such simulations are computationally expensive. In their work, the Kelvin force model has been used to calculate interactions between different SPH particles. The application of this model results in a force pointing outwards with respect to the fluid surface causing significant levitation problems. This drawback limits the application of more advanced and efficient SPH frameworks such as divergence-free SPH (DFSPH) or implicit incompressible SPH (IISPH). In this contribution, we propose a current loop magnetic force model which enables the fast macroscopic simulation of ferrofluids. Our new force model results in a force term pointing inwards allowing for more stable and fast simulations of ferrofluids using DFSPH and IISPH.

Conjugated Polymers as a New Class of Dual-Mode Matrices for MALDI Mass Spectrometry and Imaging.

Matrix-assisted laser desorption/ionization mass spectrometry (MALDI MS) and MALDI MS imaging are ubiquitous analytical methods in medical, pharmaceutical, biological, and environmental research. Currently, there is a strong interest in the investigation of low molecular weight compounds (LMWCs), especially to trace and understand metabolic pathways, requiring the development of new matrix systems that have favorable optical properties and a high ionization efficiency and that are MALDI silent in the LMWC area. In this paper, five conjugated polymers, poly{[ N, N'-bis(2-octyldodecyl)-naphtalene-1,4,5,8-bis(dicarboximide)-2,6-diyl]- alt-5,5'(2,2'-bithiophene)} (PNDI(T2)), poly(3-dodecylthiophene-2,5-diyl) (P3DDT), poly{[2,3-bis(3-octyloxyphenyl)quinoxaline-5,8-diyl]- alt-(thiophene-2,5-diyl)} (PTQ1), poly{[ N, N'-bis(2-octyldodecyl)-isoindigo-5,5'-diyl] -alt-5,5'(2,2'-bithiophene)} (PII(T2)), and poly(9,9-di- n-octylfluorenyl-2,7-diyl) (P9OFl) are investigated as matrices. The polymers have a strong optical absorption, are solution processable, and can be coated into thin films, allowing a vast reduction in the amount of matrix used. All investigated polymers function as matrices in both positive and negative mode MALDI, classifying them as rare dual-mode matrices, and show a very good analyte ionization ability in both modes. PNDI(T2), P3DDT, PTQ1, and PII(T2) are MALDI silent in the full measurement range (> m/ z = 150k), except at high laser intensities. In MALDI MS experiments of single analytes and a complex biological sample, the performance of the polymers was found to be as good as two commonly used matrices (2,5-DHB for positive and 9AA for negative mode measurements). The detection limit of two standard analytes was determined as being below 164 pmol for reserpine and below 245 pmol for cholic acid. Additionally P3DDT was used successfully in first MALDI MS imaging experiments allowing the visualization of the tissue morphology of rat brain sections.