DualNetGO: A Dual Network Model for Protein Function Prediction via Effective Feature Selection.

MOTIVATION: Protein-protein Interaction (PPI) networks are crucial for automatically annotating protein functions. As multiple PPI networks exist for the same set of proteins that capture properties from different aspects, it is a challenging task to effectively utilize these heterogeneous networks. Recently, several deep learning models have combined PPI networks from all evidence, or concatenated all graph embeddings for protein function prediction. However, the lack of a judicious selection procedure prevents the effective harness of information from different PPI networks, as these networks vary in densities, structures, and noise levels. Consequently, combining protein features indiscriminately could increase the noise level, leading to decreased model performance. RESULTS: We develop DualNetGO, a dual network model comprised of a classifier and a selector, to predict protein functions by effectively selecting features from different sources including graph embeddings of PPI networks, protein domain and subcellular location information. Evaluation of DualNetGO on human and mouse datasets in comparison with other network-based models show at least 4.5%, 6.2% and 14.2% improvement on Fmax in BP, MF and CC Gene Ontology categories respectively for human, and 3.3%, 10.6% and 7.7% improvement on Fmax for mouse. We demonstrate the generalization capability of our model by training and testing on the CAFA3 data, and show its versatility by incorporating Esm2 embeddings. We further show that our model is insensitive to the choice of graph embedding method and is time- and memory-saving. These results demonstrate that combining a subset of features including PPI networks and protein attributes selected by our model is more effective in utilizing PPI network information than only using one kind of or concatenating graph embeddings from all kinds of PPI networks. AVAILABILITY AND IMPLEMENTATION: The source code of DualNetGO and some of the experiment data are available at: https://github.com/georgedashen/DualNetGO. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

[Synergistic effect of acupuncture and moxibustion on expression of renal Fas and FasL mRNAs and proteins in contrast-induced nephropathy diabetic rats].

OBJECTIVE: To observe the effect of acupuncture and moxibustion on renal Fas and Fas ligand (FasL) expression in contrast-induced nephropathy (CIN) diabetic rats, so as to explore its protective mechanisms. METHODS: A total of 40 diabetic male SD rats were randomly divided into control, model, acupuncture, moxibustion and combined treatment (acupuncture combined with moxibustion) groups (n＝8 rats in each group). The diabetic model was established by i．p. injection of Streptozotocin (60 mg/kg), followed by regular raising for 9 weeks. Then, the manual acupuncture, moxibustion or acupuncture plus moxibustion at "Sanyinjiao"(SP6) "Shenshu"(BL23) and "Pishu"(BL20) was given to the diabetic rats, once daily for 7 days. On the 8th day, these diabetic rats received i．p. injection of compound Meglumine Diatrizoate Injection (10 mL/kg) for establishing CIN model. Then, the venous blood specimens were collected for detecting contents of serum urine nitrogen (BUN, with urease method) and creatinine (Scr, with picric acid method). The kindey tissues were exteriorized for detecting nitricoxide synthase (NOS, with oxidation-reduction method), malondialdehyde (MDA, with thiobarbituric acid method), superoxide dismutase (SOD, with xanthine oxidase technique) and total antioxidant capacity (T-AOC, with colorimetric technique). The expression levels of Fas and FasL in the kidney tissues were detected by quantitative real-time PCR and Western blot separately. The histopathological and ultrastructural changes of the kidney were observed under microscope (after H．E. staining) and transmission electron microscope, respectively. RESULTS: Following modeling and compared with the control group, the BUN, Scr and renal MDA contents, and Fas and FasL mRNA and protein expression levels were significantly increased (P<0.01, P<0.05), and renal NOS, SOD and T-AOC levels considerably decreased in the model group (P<0.05). After the interventions and compared with the model group, serum BUN and Scr contents and renal Fas mRNA and protein expression levels in both acupuncture group and combined treatment group, BUN in the moxibustion group, and renal MDA content and FasL mRNA and protein expression levels in the 3 treatment groups were obviously down-regulated (P<0.05，P<0.01), while renal NOS in the 3 treatment groups, and SOD and T-AOC levels in the combined treatment group were considerably up-regulated (P<0.01，P<0.05). The effects of the combined treatment were significantly superior to those of simple acupuncure and simple moxibustion in down-regulating serum BUN contents, and Fas and FasL mRNA and protein expression levels, and in up-regulating renal NOS, SOD and T-AOC activity (P<0.05). No significant differences were found between the acupuncture and moxibustion groups in down-regulating serum BUN, Scr and renal MDA contents, and Fas and FasL mRNA and protein levels, as well as in up-regulating NOS content (P>0.05). H．E. staining showed lobular changes of glomeruli, many local necrosis and vascular degeneration in the model group, and transmission electron microscope displayed thickened basement membrane of the glomerular blood loop, many vacuoles in the cytoplasm of renal tubular epithelial cells, severely swollen mitochondria with disordered, broken or disappeared cristae, and some apoptotic cells in the model group, which were relatively milder in the combined treatment group. CONCLUSION: Joint application of acupuncture and moxibustion can reduce the oxidative stress and renal injury in CIN diabetic rats, which may be associated with its function in down-regulating the expression of renal Fas and FasL genes and proteins. Acupuncture and moxibustion has a synergistic effect.