Chainsaw: protein domain segmentation with fully convolutional neural networks.

MOTIVATION: Protein domains are fundamental units of protein structure and play a pivotal role in understanding folding, function, evolution, and design. The advent of accurate structure prediction techniques has resulted in an influx of new structural data, making the partitioning of these structures into domains essential for inferring evolutionary relationships and functional classification. RESULTS: This article presents Chainsaw, a supervised learning approach to domain parsing that achieves accuracy that surpasses current state-of-the-art methods. Chainsaw uses a fully convolutional neural network which is trained to predict the probability that each pair of residues is in the same domain. Domain predictions are then derived from these pairwise predictions using an algorithm that searches for the most likely assignment of residues to domains given the set of pairwise co-membership probabilities. Chainsaw matches CATH domain annotations in 78% of protein domains versus 72% for the next closest method. When predicting on AlphaFold models, expert human evaluators were twice as likely to prefer Chainsaw's predictions versus the next best method. AVAILABILITY AND IMPLEMENTATION: github.com/JudeWells/Chainsaw.

Enhancing missense variant pathogenicity prediction with protein language models using VariPred.

Computational approaches for predicting the pathogenicity of genetic variants have advanced in recent years. These methods enable researchers to determine the possible clinical impact of rare and novel variants. Historically these prediction methods used hand-crafted features based on structural, evolutionary, or physiochemical properties of the variant. In this study we propose a novel framework that leverages the power of pre-trained protein language models to predict variant pathogenicity. We show that our approach VariPred (Variant impact Predictor) outperforms current state-of-the-art methods by using an end-to-end model that only requires the protein sequence as input. Using one of the best-performing protein language models (ESM-1b), we establish a robust classifier that requires no calculation of structural features or multiple sequence alignments. We compare the performance of VariPred with other representative models including 3Cnet, Polyphen-2, REVEL, MetaLR, FATHMM and ESM variant. VariPred performs as well as, or in most cases better than these other predictors using six variant impact prediction benchmarks despite requiring only sequence data and no pre-processing of the data.

Complex care needs and devolution in Greater Manchester: a pilot study to explore social care innovation in newly integrated service arrangements for older people.

Objective The aim of this study was to describe emergent approaches to integrated care for older people with complex care needs and investigate the viability of measuring integrated care. Methods A case study approach was used. Sites were recruited following discussion with senior staff in health and social care agencies. Service arrangements were categorised using a framework developed by the researchers. To investigate joint working within the sites, the development model for integrated care was adapted and administered to the manager of each service. Data were collected in 2018. Results Six case study sites were recruited illustrating adult social care services partnerships in services for older people with home care providers, mental health and community nursing services. Most were established in 2018. Service arrangements were characterised by joint assessment and informal face-to-face discussions between staff. The development of an infrastructure to promote partnership working was evident between adult social care and each of the other services and most developed with home care providers. There was little evidence of a sequential approach to the development of integrated working practices. Conclusion Components of partnerships promoting integrated care have been highlighted and understanding of the complexity of measuring integrated care enhanced. Means of information sharing and work force development require further consideration. What is known about the topic? The devolution of health and social care arrangements in Greater Manchester has aroused considerable interest in much wider arenas. Necessarily much of the focus in available material has been upon strategic development, analysis of broader trends and mechanisms and a concern with changes in the healthcare system. What does this paper add? The findings from this study will enable emerging approaches to be described and codified, and permit the specific social care contribution to the new arrangements to be discerned. The findings are relevant beyond the immediate context of Greater Manchester to wider integrated care. The evidence can be used by commissioners and services, providing a sound basis for further work as service systems develop. What are the implications for practitioners? This research is important because it is one of the first pieces of work to examine the new integrated care arrangements in Greater Manchester. By providing guidance to promote evidence-based practice, this study contributes to service development in Greater Manchester and the achievement of the broad national service objectives of improving user and carer experiences and ensuring value for money.