We address the problem of generating novel molecules with desired interaction properties as a multi-objective optimization problem. Interaction binding models are learned from binding data using ...graph convolution networks (GCNs). Since the experimentally obtained property scores are recognised as having potentially gross errors, we adopted a robust loss for the model. Combinations of these terms, including drug likeness and synthetic accessibility, are then optimized using reinforcement learning based on a graph convolution policy approach. Some of the molecules generated, while legitimate chemically, can have excellent drug-likeness scores but appear unusual. We provide an example based on the binding potency of small molecules to dopamine transporters. We extend our method successfully to use a multi-objective reward function, in this case for generating novel molecules that bind with dopamine transporters but not with those for norepinephrine. Our method should be generally applicable to the generation in silico of molecules with desirable properties.
Recent research in multilingual language models (LM) has demonstrated their ability to effectively handle multiple languages in a single model. This holds promise for low web-resource languages (LRL) ...as multilingual models can enable transfer of supervision from high resource languages to LRLs. However, incorporating a new language in an LM still remains a challenge, particularly for languages with limited corpora and in unseen scripts. In this paper we argue that relatedness among languages in a language family may be exploited to overcome some of the corpora limitations of LRLs, and propose RelateLM. We focus on Indian languages, and exploit relatedness along two dimensions: (1) script (since many Indic scripts originated from the Brahmic script), and (2) sentence structure. RelateLM uses transliteration to convert the unseen script of limited LRL text into the script of a Related Prominent Language (RPL) (Hindi in our case). While exploiting similar sentence structures, RelateLM utilizes readily available bilingual dictionaries to pseudo translate RPL text into LRL corpora. Experiments on multiple real-world benchmark datasets provide validation to our hypothesis that using a related language as pivot, along with transliteration and pseudo translation based data augmentation, can be an effective way to adapt LMs for LRLs, rather than direct training or pivoting through English.
The segmentation of neck muscles is useful for the diagnoses and planning of medical interventions for neck pain-related conditions such as whiplash and cervical dystonia. Neck muscles are tightly ...grouped, have similar appearance to each other and display large anatomical variability between subjects. They also exhibit low contrast with background organs in magnetic resonance (MR) images. These characteristics make the segmentation of neck muscles a challenging task. Due to the significant success of the U-Net architecture for deep learning-based segmentation, numerous versions of this approach have emerged for the task of medical image segmentation. This paper presents an evaluation of 10 U-Net CNN approaches, 6 direct (U-Net, CRF-Unet, A-Unet, MFP-Unet, R2Unet and U-Net++) and 4 modified (R2A-Unet, R2A-Unet++, PMS-Unet and MS-Unet). The modifications are inspired by recent multi-scale and multi-stream techniques for deep learning algorithms. T1 weighted axial MR images of the neck, at the distal end of the C3 vertebrae, from 45 subjects with real-time data augmentation were used in our evaluation of neck muscle segmentation approaches. The analysis of our numerical results indicates that the R2Unet architecture achieves the best accuracy.