Identifying Membrane Protein Types Based on Lifelong Learning With Dynamically Scalable Networks

Identification of membrane protein types via deep residual hypergraph neural network

A membrane protein's functions are significantly associated with its type, so it is crucial to identify the types of membrane proteins. Conventional computational methods for identifying the species of membrane proteins tend to ignore two issues: High-order correlation among membrane proteins and the scenarios of multi-modal representations of membrane proteins, which leads to information loss. To tackle those two issues, we proposed a deep residual hypergraph neural network (DRHGNN), which enhances the hypergraph neural network (HGNN) with initial residual and identity mapping in this paper. We carried out extensive experiments on four benchmark datasets of membrane proteins. In the meantime, we compared the DRHGNN with recently developed advanced methods. Experimental results showed the better performance of DRHGNN on the membrane protein classification task on four datasets. Experiments also showed that DRHGNN can handle the over-smoothing issue with the increase of the number of model layers compared with HGNN. The code is available at https://github.com/yunfighting/Identification-of-Membrane-Protein-Types-via-deep-residual-hypergraph-neural-network.

Molecular remodeling of adipose tissue is associated with metabolic recovery after weight loss surgery

BACKGROUND

Bariatric surgery is an effective therapy for individuals with severe obesity to achieve sustainable weight loss and to reduce comorbidities. Examining the molecular signature of subcutaneous adipose tissue (SAT) following different types of bariatric surgery may help in gaining further insight into their distinct metabolic impact.

RESULTS

Subjects undergoing biliopancreatic diversion with duodenal switch (BPD-DS) showed a significantly higher percentage of total weight loss than those undergoing gastric bypass or sleeve gastrectomy (RYGB + SG) (41.7 ± 4.6 vs 28.2 ± 6.8%; p = 0.00005). Individuals losing more weight were also significantly more prone to achieve both type 2 diabetes and dyslipidemia remission (OR = 0.75; 95%CI = 0.51-0.91; p = 0.03). Whole transcriptome and methylome profiling showed that bariatric surgery induced a profound molecular remodeling of SAT at 12 months postoperative, mainly through gene down-regulation and hypermethylation. The extent of changes observed was greater following BPD-DS, with 61.1% and 49.8% of up- and down-regulated genes, as well as 85.7% and 70.4% of hyper- and hypomethylated genes being exclusive to this procedure, and mostly associated with a marked decrease of immune and inflammatory responses. Weight loss was strongly associated with genes being simultaneously differentially expressed and methylated in BPD-DS, with the strongest association being observed for GPD1L (r² = 0.83; p = 1.4 × 10^-6).

CONCLUSIONS

Present findings point to the greater SAT molecular remodeling following BPD-DS as potentially linked with higher metabolic remission rates. These results will contribute to a better understanding of the metabolic pathways involved in the response to bariatric surgery and will eventually lead to the development of gene targets for the treatment of obesity. Trial registration ClinicalTrials.gov NCT02390973.