Mesoporous Silicon with Strontium-Powered Poly(Lactic-Co-Glycolic acid)/Gelatin-Based Dressings Facilitate Skin Tissue Repair.

Purpose: Functional inorganic nanomaterials (NMs) are widely exploited as bioactive materials and drug depots. The lack of a stable form of application of NMs at the site of skin injury, may impede the removal of the debridement, elevate pH, induce tissue toxicity, and limit their use in skin repair. This necessitates the advent of innovative wound dressings that overcome the above limitations. The overarching objective of this study was to exploit strontium-doped mesoporous silicon particles (PSiSr) to impart multifunctionality to poly(lactic-co-glycolic acid)/gelatin (PG)-based fibrous dressings (PG@PSiSr) for excisional wound management. Methods: Mesoporous silicon particles (PSi) and PSiSr were synthesized using a chemo-synthetic approach. Both PSi and PSiSr were incorporated into PG fibers using electrospinning. A series of structure, morphology, pore size distribution, and cumulative pH studies on the PG@PSi and PG@PSiSr membranes were performed. Cytocompatibility, hemocompatibility, transwell migration, scratch wound healing, and delineated angiogenic properties of these composite dressings were tested in vitro. The biocompatibility of composite dressings in vivo was assessed by a subcutaneous implantation model of rats, while their potential for wound healing was discerned by implantation in a full-thickness excisional defect model of rats. Results: The PG@PSiSr membranes can afford the sustained release of silicon ions (Si4+) and strontium ions (Sr2+) for up to 192 h as well as remarkably promote human umbilical vein endothelial cells (HUVECs) and NIH-3T3 fibroblasts migration. The PG@PSiSr membranes also showed better cytocompatibility, hemocompatibility, and significant formation of tubule-like networks of HUVECs in vitro. Moreover, PG@PSiSr membranes also facilitated the infiltration of host cells and promoted the deposition of collagen while reducing the accumulation of inflammatory cells in a subcutaneous implantation model in rats as assessed for up to day 14. Further evaluation of membranes transplanted in a full-thickness excisional wound model in rats showed rapid wound closure (PG@SiSr vs control, 96.1% vs 71.7%), re-epithelialization, and less inflammatory response alongside skin appendages formation (eg, blood vessels, glands, hair follicles, etc.). Conclusion: To sum up, we successfully fabricated PSiSr particles and prepared PG@PSiSr dressings using electrospinning. The PSiSr-mediated release of therapeutic ions, such as Si4+ and Sr2+, may improve the functionality of PLGA/Gel dressings for an effective wound repair, which may also have implications for the other soft tissue repair disciplines.

Widely Targeted Metabolomic Analysis Reveals the Improvement in Panax notoginseng Triterpenoids Triggered by Arbuscular Mycorrhizal Fungi via UPLC-ESI-MS/MS.

Panax notoginseng is a highly valued perennial medicinal herb in China and is widely used in clinical treatments. The main purpose of this study was to elucidate the changes in the composition of P. notoginseng saponins (PNSs), which are the main bioactive substances, triggered by arbuscular mycorrhizal fungi (AMF) via ultrahigh-performance liquid chromatography-electrospray ionization-tandem mass spectrometry (UPLC-ESI-MS/MS). A total of 202 putative terpenoid metabolites were detected, of which 150 triterpene glycosides were identified, accounting for 74.26% of the total. Correlation analysis, principal component analysis (PCA) and orthogonal partial least squares discriminant analysis (OPLS-DA) of the metabolites revealed that the samples treated with AMF (group Ce) could be clearly separated from the CK samples. In total, 49 differential terpene metabolites were identified between the Ce and CK groups, of which 38 and 11 metabolites were upregulated and downregulated, respectively, and most of the upregulated differentially abundant metabolites were mainly triterpene glycosides. The relative abundances of the two major notoginsenosides (MNs), ginsenosides Rd and Re, and 13 rare notoginsenosides (RNs), significantly increased. The differential saponins, especially RNs, were more easily clustered into one branch and had a high positive correlation. It could be concluded that the biosynthesis and accumulation of some RNs share the same pathways as those triggered by AMF. This study provides a new way to obtain more notoginsenoside resources, particularly RNs, and sheds new light on the scientization and rationalization of the use of AMF agents in the ecological planting of medicinal plants.

Inhibiting NF-κB inducing kinase improved the motor performance of ALS animal model.

BACKGROUND: Amyotrophic lateral sclerosis (ALS) is a typical neurodegenerative disorder typically characterized by inflammation activation. However, the relationship between non-canonical NF-κB (ncNF-κB) pathway activation and ALS progression is not clear. METHODS: We tested the ncNF-κB pathway in the ALS animal model including hSOD1-G93A transgenic mice and TBK1 deletion mice.We treated age-matched SOD1-G93A mice with B022 (a NIK inhibitor) to investigate the role of NIK in the ALS animal model. We also established a new mice model by crossing SOD1-G93A mice with NIK+/- mice to further evaluate the interrelationship between the NIK and the disease progression in ALS animal model. RESULTS: In this study, we found the ncNF-κB pathway was activated in SOD1-G93A animal model and TBK1 deletion model. Inhibition of NIK activity by small molecule B022 significantly improved the motor performance of the ALS animal model. However, NIK deletion enhanced the mutant SOD1 toxicity by inflammatory infiltration. CONCLUSION: TBK1 deletion and mutant SOD1 shared the common pathological feature possibly via effects on NIK activation and inhibitor of NIK could be a novel strategy for treating ALS.

Improvement of Panax notoginseng saponin accumulation triggered by methyl jasmonate under arbuscular mycorrhizal fungi.

Panax notoginseng is a highly valued perennial medicinal herb plant in Yunnan Province, China, and the taproots are the main medicinal parts that are rich in active substances of P. notoginseng saponins. The main purpose of this study is to uncover the physiological and molecular mechanism of Panax notoginseng saponin accumulation triggered by methyl jasmonate (MeJA) under arbuscular mycorrhizal fungi (AMF) by determining physiological indices, high-throughput sequencing and correlation analysis. Physiological results showed that the biomass and saponin contents of P. notoginseng, the concentrations of jasmonic acids (JAs) and the key enzyme activities involved in notoginsenoside biosynthesis significantly increased under AMF or MeJA, but the interactive treatment of AMF and MeJA weakened the effect of AMF, suggesting that a high concentration of endogenous JA have inhibitory effect. Transcriptome sequencing results indicated that differential expressed genes (DEGs) involved in notoginsenoside and JA biosynthesis were significantly enriched in response to AMF induction, e.g., upregulated genes of diphosphocytidyl-2-C-methyl-d-erythritol kinases (ISPEs), cytochrome P450 monooxygenases (CYP450s)_and glycosyltransferases (GTs), while treatments AMF-MeJA and salicylhydroxamic acid (SHAM) decreased the abundance of these DEGs. Interestingly, a high correlation presented between any two of saponin contents, key enzyme activities and expression levels of DEGs. Taken together, the inoculation of AMF can improve the growth and saponin accumulation of P. notoginseng by strengthening the activities of key enzymes and the expression levels of encoding genes, in which the JA regulatory pathway is a key link. This study provides references for implementing ecological planting of P. notoginseng, improving saponin accumulation and illustrating the biosynthesis mechanism.

MERITS: a web-based integrated Mycobacterial PE/PPE protein database.

Motivation: PE/PPE proteins, highly abundant in the Mycobacterium genome, play a vital role in virulence and immune modulation. Understanding their functions is key to comprehending the internal mechanisms of Mycobacterium. However, a lack of dedicated resources has limited research into PE/PPE proteins. Results: Addressing this gap, we introduce MycobactERIal PE/PPE proTeinS (MERITS), a comprehensive 3D structure database specifically designed for PE/PPE proteins. MERITS hosts 22 353 non-redundant PE/PPE proteins, encompassing details like physicochemical properties, subcellular localization, post-translational modification sites, protein functions, and measures of antigenicity, toxicity, and allergenicity. MERITS also includes data on their secondary and tertiary structure, along with other relevant biological information. MERITS is designed to be user-friendly, offering interactive search and data browsing features to aid researchers in exploring the potential functions of PE/PPE proteins. MERITS is expected to become a crucial resource in the field, aiding in developing new diagnostics and vaccines by elucidating the sequence-structure-functional relationships of PE/PPE proteins. Availability and implementation: MERITS is freely accessible at http://merits.unimelb-biotools.cloud.edu.au/.

Catalysts of Healing: A Symphony of Synthesis and Clinical Artistry in Small-Molecule Agents for Breast Cancer Alleviation.

Breast cancer, characterized by its molecular intricacy, has witnessed a surge in targeted therapeutics owing to the rise of small-molecule drugs. These entities, derived from cutting-edge synthetic routes, often encompassing multistage reactions and chiral synthesis, target a spectrum of oncogenic pathways. Their mechanisms of action range from modulating hormone receptor signaling and inhibiting kinase activity, to impeding DNA damage repair mechanisms. Clinical applications of these drugs have resulted in enhanced patient survival rates, reduction in disease recurrence, and improved overall therapeutic indices. Notably, certain molecules have showcased efficacy in drug-resistant breast cancer phenotypes, highlighting their potential in addressing treatment challenges. The evolution and approval of small-molecule drugs have ushered in a new era for breast cancer therapeutics. Their tailored synthetic pathways and defined mechanisms of action have augmented the precision and efficacy of treatment regimens, paving the way for improved patient outcomes in the face of this pervasive malignancy. The present review embarks on a detailed exploration of small-molecule drugs that have secured regulatory approval for breast cancer treatment, emphasizing their clinical applications, synthetic pathways, and distinct mechanisms of action.

Characterization of two Bursicon genes and their association with wing development in the brown citrus aphid, Aphis citricidus.

The tanning hormone, Bursicon, is a neuropeptide secreted by the insect nervous system that functions as a heterodimer composed of Burs-α and Burs-ß subunits. It plays a critical role in the processes of cuticle tanning and wing expansion in insects. In this study, we successfully identified the AcBurs-α and AcBurs-ß genes in Aphis citricidus. The open reading frames of AcBurs-α and AcBurs-ß were 480 and 417 bp in length, respectively. Both AcBurs-α and AcBurs-ß exhibited 11 conserved cysteine residues. AcBurs-α and AcBurs-ß were expressed during all developmental stages of A. citricidus and showed high expression levels in the winged aphids. To investigate the potential role of AcBurs-α and AcBurs-ß in wing development, we employed RNA interference (RNAi) techniques. With the efficient silencing of AcBurs-α (44.90%) and AcBurs-ß (52.31%), malformed wings were induced in aphids. The proportions of malformed wings were 22.50%, 25.84%, and 38.34% in dsAcBurs-α-, dsAcBur-ß-, and dsAcBurs-α + dsAcBur-ß-treated groups, respectively. Moreover, feeding protein kinase A inhibitors (H-89) also increased the proportion of malformed wings to 30.00%. Feeding both double-stranded RNA and inhibitors (H-89) significantly downregulated the wing development-related genes nubbin, vestigial, notch and spalt major. Silence of vestigial through RNAi also led to malformed wings. Meanwhile, the exogenous application of 3 hormones that influence wing development did not affect the expression level of AcBursicon genes. These findings indicate that AcBursicon genes plays a crucial role in wing development in A. citricidus; therefore, it represents a potential molecular target for the control of this pest through RNAi-based approaches.

Multidrug resistance of Botrytis cinerea associated with its adaptation to plant secondary metabolites.

Fungicides are an effective way to control gray mold of grapes, but the pathogen Botrytis cinerea can develop resistance, overcoming the effectiveness of a fungicide that is repeatedly applied. More importantly, the emergence of multidrug resistance (MDR) in the field, where multiple fungicides with different modes of action simultaneously lose their efficacies, is a significant concern. MDR is associated with ATP-binding cassette (ABC) transporters of the pathogen, and certain plant secondary metabolites (PSMs) stimulate the upregulation of ABC transporters, we hypothesized that the pathogen's preadaptation to PSMs might contribute to MDR development. To test this in B. cinerea, ten PSMs, namely, resveratrol, reserpine, chalcone, flavanone, eugenol, farnesol, anethene, camptothecin, salicylic acid, and psoralen, were selected based on their association with ABC transporters involved in fungicide resistance. B. cinerea strain B05.10 was continuously transferred for 15 generations on potato dextrose agar amended with a PSM (PDAP), and sensitivities to PSMs and fungicides were examined on the 5th, 10th, and 15th generations. RNA was extracted from B. cinerea from the selected generations. After 15 generations of culture transfers, an up-regulation was observed in the expression of ABC transporter-encoding genes BcatrB, BcatrD, and BcatrK using quantitative polymerase chain reaction (qPCR). This upregulation was found to contribute to MDR of B. cinerea against two or more fungicides, among azoxystrobin, boscalid, fludioxonil, difenoconazole, prochloraz, and pyrimethanil. This finding was confirmed through genetic transformation. The decreased sensitivity of B. cinerea to fungicides was confirmed as a subsequent MDR phenotype after exposure to camptothecin, flavanone, and resveratrol. Besides, transcriptome analysis also revealed the upregulation of transcription factors related to ABC expression following resveratrol exposure. This suggests that PSMs contributed to inducing preadaptation of B. cinerea, leading to subsequent MDR.IMPORTANCEThe emergence of MDR in plant pathogens is a threat to plant disease management and leads to the use of excessive fungicides. Botrytis cinerea is of particular concern because its MDR has widely emerged in the field. Understanding its genesis is the first step for controlling MDR. In this study, the contribution of PSMs to MDR has been examined. Effective management of this pathogen in agroecosystems relies on a better understanding of how it copes with phytochemicals or fungicides.

Factors associated with eating rate: a systematic review and narrative synthesis informed by socio-ecological model.

Accumulating evidence shows associations between rapid eating and overweight. Modifying eating rate might be a potential weight management strategy without imposing additional dietary restrictions. A comprehensive understanding of factors associated with eating speed will help with designing effective interventions. The aim of this review was to synthesise the current state of knowledge on the factors associated with eating rate. The socio-ecological model (SEM) was utilised to scaffold the identified factors. A comprehensive literature search of eleven databases was conducted to identify factors associated with eating rate. The 104 studies that met the inclusion criteria were heterogeneous in design and methods of eating rate measurement. We identified thirty-nine factors that were independently linked to eating speed and mapped them onto the individual, social and environmental levels of the SEM. The majority of the reported factors pertained to the individual characteristics (n = 20) including demographics, cognitive/psychological factors and habitual food oral processing behaviours. Social factors (n = 11) included eating companions, social and cultural norms, and family structure. Environmental factors (n = 8) included food texture and presentation, methods of consumption or background sounds. Measures of body weight, food form and characteristics, food oral processing behaviours and gender, age and ethnicity were the most researched and consistent factors associated with eating rate. A number of other novel and underresearched factors emerged, but these require replication and further research. We highlight directions for further research in this space and potential evidence-based candidates for interventions targeting eating rate.

High expression of NF-κB inducing kinase in the bulge region of hair follicle induces tumor.

The bulge region, a reservoir of multipotent stem cells, is possibly responsible for tumorigenesis. NF-κB-inducing kinase (NIK) is a kinase involved in the activation of the noncanonical NF-κB pathway and exhibits positive staining in tumor cells. However, whether high expression of NIK can result in tumorigenesis has not been reported in published papers. By establishing Nik-coe (Nik-stopF/F crossed with Chat-cre) and Nik-soe (Nik-stopF/F crossed with Sox9-cre) mice, we found that overexpression of Nik in the bulge region of hair follicles induced hair follicle loss and tumorigenesis. Furthermore, RNA sequencing, proteomic and phosphopeptide analyses revealed that multiple cancer pathways are involved in tumor formation. Taken together, these findings indicate that constitutive activation of Nik in the bulge region induces tumorigenesis.

Comprehensive Analysis of 34 Edible Flowers by the Determination of Nutritional Composition and Antioxidant Capacity Planted in Yunnan Province China.

The main purpose of this study was to reveal the nutritional value and antioxidant activity of 34 edible flowers that grew in Yunnan Province, China, through a comprehensive assessment of their nutritional composition and antioxidant indices. The results showed that sample A3 of Asteraceae flowers had the highest total flavonoid content, with a value of 8.53%, and the maximum contents of vitamin C and reducing sugars were from Rosaceae sample R1 and Gentianaceae sample G3, with values of 143.80 mg/100 g and 7.82%, respectively. Samples R2 and R3 of Rosaceae were the top two flowers in terms of comprehensive nutritional quality. In addition, the antioxidant capacity of Rosaceae samples was evidently better than that of three others, in which Sample R1 had the maximum values in hydroxyl radical (·OH) scavenging and superoxide anion radical (·O2-) scavenging rates, and samples R2 and R3 showed a high total antioxidant capacity and 2,2-diphenyl-1-pyridylhydrazine (DPPH) scavenging rate, respectively. Taken together, there were significant differences in the nutrient contents and antioxidant properties of these 34 flowers, and the comprehensive quality of Rosaceae samples was generally better than the other three families. This study provides references for 34 edible flowers to be used as dietary supplements and important sources of natural antioxidants.

Digerati - A multipath parallel hybrid deep learning framework for the identification of mycobacterial PE/PPE proteins.

The genome of Mycobacterium tuberculosis contains a relatively high percentage (10%) of genes that are poorly characterised because of their highly repetitive nature and high GC content. Some of these genes encode proteins of the PE/PPE family, which are thought to be involved in host-pathogen interactions, virulence, and disease pathogenicity. Members of this family are genetically divergent and challenging to both identify and classify using conventional computational tools. Thus, advanced in silico methods are needed to identify proteins of this family for subsequent functional annotation efficiently. In this study, we developed the first deep learning-based approach, termed Digerati, for the rapid and accurate identification of PE and PPE family proteins. Digerati was built upon a multipath parallel hybrid deep learning framework, which equips multi-layer convolutional neural networks with bidirectional, long short-term memory, equipped with a self-attention module to effectively learn the higher-order feature representations of PE/PPE proteins. Empirical studies demonstrated that Digerati achieved a significantly better performance (∼18-20%) than alignment-based approaches, including BLASTP, PHMMER, and HHsuite, in both prediction accuracy and speed. Digerati is anticipated to facilitate community-wide efforts to conduct high-throughput identification and analysis of PE/PPE family members. The webserver and source codes of Digerati are publicly available at http://web.unimelb-bioinfortools.cloud.edu.au/Digerati/.

Long non-coding RNA DINO promotes cisplatin sensitivity in lung adenocarcinoma via the p53-Bax axis.

Background: The damage-induced non-coding (DINO) RNA is a newly identified long non-coding RNA (lncRNA) found in human cells with DNA damage. The treatment of tumors with cisplatin can induce DNA damage; however, whether the lncRNA DINO is involved in the treatment of non-small cell lung cancer (NSCLC) has not yet been elucidated. Methods: The expression of the lncRNA DINO in lung adenocarcinoma cells was detected using quantitative real-time polymerase chain reaction (qRT-PCR). The lung adenocarcinoma cell line, A549, and derived cisplatin-resistant cell line, A549R, were selected to construct cell models with lncRNA DINO overexpression or interference via lentiviral transfection. After cisplatin treatment, changes in the apoptosis rate were measured. Changes in the p53-Bax axis were detected by qRT-PCR and Western blot. Cycloheximide (CHX) interference demonstrated the stability of p53 with new protein production induced by the lncRNA DINO. The in vivo experiments involved intraperitoneal injection of nude mice with cisplatin after subcutaneous tumor formation, and the tumor diameters and weights were recorded. Immunohistochemistry and hematoxylin and eosin (H&E) staining were performed following tumor removal. Results: We found that the lncRNA DINO was significantly down-regulated in NSCLC. DINO overexpression enhanced the sensitivity of NSCLC cells to cisplatin, while DINO down-regulation decreased the sensitivity of NSCLC cells to cisplatin. Mechanistic investigation indicated that DINO enhanced the stability of p53 and mediated the activation of the p53-Bax signaling axis. Our results also demonstrated that the lncRNA DINO could partially reverse cisplatin resistance induced by silencing the p53-Bax axis, and could inhibit subcutaneous tumorigenesis in nude mice after cisplatin treatment in vivo. Conclusions: The lncRNA DINO regulates the sensitivity of lung adenocarcinoma to cisplatin by stabilizing p53 and activating the p53-Bax axis, and thus, may be a novel therapeutic target to overcome cisplatin resistance.

ATTIC is an integrated approach for predicting A-to-I RNA editing sites in three species.

A-to-I editing is the most prevalent RNA editing event, which refers to the change of adenosine (A) bases to inosine (I) bases in double-stranded RNAs. Several studies have revealed that A-to-I editing can regulate cellular processes and is associated with various human diseases. Therefore, accurate identification of A-to-I editing sites is crucial for understanding RNA-level (i.e. transcriptional) modifications and their potential roles in molecular functions. To date, various computational approaches for A-to-I editing site identification have been developed; however, their performance is still unsatisfactory and needs further improvement. In this study, we developed a novel stacked-ensemble learning model, ATTIC (A-To-I ediTing predICtor), to accurately identify A-to-I editing sites across three species, including Homo sapiens, Mus musculus and Drosophila melanogaster. We first comprehensively evaluated 37 RNA sequence-derived features combined with 14 popular machine learning algorithms. Then, we selected the optimal base models to build a series of stacked ensemble models. The final ATTIC framework was developed based on the optimal models improved by the feature selection strategy for specific species. Extensive cross-validation and independent tests illustrate that ATTIC outperforms state-of-the-art tools for predicting A-to-I editing sites. We also developed a web server for ATTIC, which is publicly available at http://web.unimelb-bioinfortools.cloud.edu.au/ATTIC/. We anticipate that ATTIC can be utilized as a useful tool to accelerate the identification of A-to-I RNA editing events and help characterize their roles in post-transcriptional regulation.

[Acupuncture for delayed sleep-wake phase disorder: a randomized controlled trial].

OBJECTIVE: To observe the clinical effect of acupuncture for delayed sleep-wake phase disorder (DSWPD). METHODS: A total of 84 patients with DSWPD were randomized into an observation group (42 cases, 2 cases dropped off) and a control group (42 cases, 3 cases dropped off). On the basis of sleep hygiene education, acupuncture was applied at Shenmai (BL 62), Zhaohai (KI 6), Hegu (LI 4), Taichong (LR 3), Zusanli (ST 36) and Sanyinjiao (SP 6) in the observation group, while placebo acupuncture was applied at the same acupoints in the control group. The treatment lasted for 8 weeks, once every other day, 3 times a week in the 1st to 4th weeks; once every 3 days, 2 times a week in the 5th to 8th weeks. Before and after treatment, the actigraphy (ACT) indexes of objective sleep (total time of stay in bed, total sleep time, sleep efficiency, the number of awakenings and the wake time after falling asleep) and plasma cortisol (CORT) level were observed; before and after treatment and in follow-up of 1, 3 months after treatment, the scores of morningness-eveningness questionnaire (MEQ), insomnia severity index (ISI), fatigue severity scale (FSS) and Epworth sleepiness scale (ESS) were observed in the two groups. RESULTS: Compared before treatment, the total sleep time was prolonged, the sleep efficiency was improved, the number of awakenings was reduced, and the wake time after falling asleep was shortened after treatment in the observation group (P<0.01, P<0.05), and those in the observation group after treatment were superior to the control group (P<0.01, P<0.05). Compared before treatment, the MEQ scores after treatment in both groups and in the follow-up of 1, 3 months after treatment in the observation group were increased (P<0.01), and the MEQ score of each time point after treatment in the observation group was higher than the control group (P<0.01). The scores of ISI, FSS and ESS after treatment, and the scores of ISI、ESS in follow-up of 1, 3 months after treatment in the observation group were decreased compared with those before treatment (P<0.01, P<0.05), and in the observation group, the scores of ISI, FSS and ESS of each time point after treatment were lower than those in the control group (P<0.01, P<0.05). After treatment, the plasma CORT level in the observation group was decreased compared with that before treatment and that in the control group (P<0.01, P<0.05). CONCLUSION: Acupuncture can improve the sleep and wake phase of patients with DSWPD, improve sleep quality and daytime function, and its mechanism may be related to the down-regulation of plasma CORT level.

Targeting tumor heterogeneity: multiplex-detection-based multiple instance learning for whole slide image classification.

MOTIVATION: Multiple instance learning (MIL) is a powerful technique to classify whole slide images (WSIs) for diagnostic pathology. The key challenge of MIL on WSI classification is to discover the critical instances that trigger the bag label. However, tumor heterogeneity significantly hinders the algorithm's performance. RESULTS: Here, we propose a novel multiplex-detection-based multiple instance learning (MDMIL) which targets tumor heterogeneity by multiplex detection strategy and feature constraints among samples. Specifically, the internal query generated after the probability distribution analysis and the variational query optimized throughout the training process are utilized to detect potential instances in the form of internal and external assistance, respectively. The multiplex detection strategy significantly improves the instance-mining capacity of the deep neural network. Meanwhile, a memory-based contrastive loss is proposed to reach consistency on various phenotypes in the feature space. The novel network and loss function jointly achieve high robustness towards tumor heterogeneity. We conduct experiments on three computational pathology datasets, e.g. CAMELYON16, TCGA-NSCLC, and TCGA-RCC. Benchmarking experiments on the three datasets illustrate that our proposed MDMIL approach achieves superior performance over several existing state-of-the-art methods. AVAILABILITY AND IMPLEMENTATION: MDMIL is available for academic purposes at https://github.com/ZacharyWang-007/MDMIL.

PFresGO: an attention mechanism-based deep-learning approach for protein annotation by integrating gene ontology inter-relationships.

MOTIVATION: The rapid accumulation of high-throughput sequence data demands the development of effective and efficient data-driven computational methods to functionally annotate proteins. However, most current approaches used for functional annotation simply focus on the use of protein-level information but ignore inter-relationships among annotations. RESULTS: Here, we established PFresGO, an attention-based deep-learning approach that incorporates hierarchical structures in Gene Ontology (GO) graphs and advances in natural language processing algorithms for the functional annotation of proteins. PFresGO employs a self-attention operation to capture the inter-relationships of GO terms, updates its embedding accordingly and uses a cross-attention operation to project protein representations and GO embedding into a common latent space to identify global protein sequence patterns and local functional residues. We demonstrate that PFresGO consistently achieves superior performance across GO categories when compared with 'state-of-the-art' methods. Importantly, we show that PFresGO can identify functionally important residues in protein sequences by assessing the distribution of attention weightings. PFresGO should serve as an effective tool for the accurate functional annotation of proteins and functional domains within proteins. AVAILABILITY AND IMPLEMENTATION: PFresGO is available for academic purposes at https://github.com/BioColLab/PFresGO. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.