Multimodal Drug Target Binding Affinity Prediction using Graph Local Substructure.

Predicting the binding affinity of drug target is essential to reduce drug development costs and cycles. Recently, several deep learning-based methods have been proposed to utilize the structural or sequential information of drugs and targets to predict the drug-target binding affinity (DTA). However, methods that rely solely on sequence features do not consider hydrogen atom data, which may result in information loss. Graph-based methods may contain information that is not directly related to the prediction process. Additionally, the lack of structured division can limit the representation of characteristics. To address these issues, we propose a multimodal DTA prediction model using graph local substructures, called MLSDTA. This model comprehensively integrates the graph and sequence modal information from drugs and targets, achieving multimodal fusion through a cross-attention approach for multimodal features. Additionally, adaptive structure aware pooling is applied to generate graphs containing local substructural information. The model also utilizes the DropNode strategy to enhance the distinctions between different molecules. Experiments on two benchmark datasets have shown that MLSDTA outperforms current state-of-the-art models, demonstrating the feasibility of MLSDTA.

Multimodal contrastive representation learning for drug-target binding affinity prediction.

In the biomedical field, the efficacy of most drugs is demonstrated by their interactions with targets, meanwhile, accurate prediction of the strength of drug-target binding is extremely important for drug development efforts. Traditional bioassay-based drug-target binding affinity (DTA) prediction methods cannot meet the needs of drug R&D in the era of big data. Recent years we have witnessed significant success on deep learning-based models for drug-target binding affinity prediction task. However, these models only considered a single modality of drug and target information, and some valuable information was not fully utilized. In fact, the information of different modalities of drug and target can complement each other, and more valuable information can be obtained by fusing the information of different modalities. In this paper, we introduce a multimodal information fusion model for DTA prediction that is called FMDTA, which fully considers drug/target information in both string and graph modalities and balances the feature representations of different modalities by a contrastive learning approach. In addition, we exploited the alignment information of drug atoms and target residues to capture the positional information of string patterns, which can extract more useful feature information in SMILES and target sequences. Experimental results on two benchmark datasets show that FMDTA outperforms the state-of-the-art model, demonstrating the feasibility and excellent feature capture capability of FMDTA. The code of FMDTA and the data are available at: https://github.com/bestdoubleLin/FMDTA.

Coexpressing the Signal Peptide of Vip3A and the Trigger Factor of Bacillus thuringiensis Enhances the Production Yield and Solubility of eGFP in Escherichia coli.

Many fusion tags have been developed to improve the expression of recombinant proteins. Besides the translocation of cargo proteins, the signal peptides (SPs) of some secretory proteins, such as the ssTorA and Iasp, have been used as an inclusion body tag (IB-tag) or the recombinant expression enhancer in the cytosol of E. coli. In this study, the approach to utilize the SP of Vip3A (Vasp) from Bacillus thuringiensis (Bt) as a fusion tag was investigated. The results showed that either the Vasp or its predicted N- (VN), H- (VH), and C-regions (VC), as well as their combinations (VNH, VNC, and VHC), were able to significantly enhance the production yield of eGFP. However, the hydrophobic region of the Vasp (VH and/or VC) made more than half of the eGFP molecules aggregated (VeGFP, VHeGFP, VCeGFP, VNHeGFP, VNCeGFP, and VHCeGFP). Interestingly, the addition of the Bt trigger factor (BtTF) led to the neutralization of the negative impact and solubilization of the fusion proteins. Therefore, the coexpression of Vasp or its derivates with the chaperone BtTF could be a novel dual-enhancement system for the production yield and solubility of recombinant proteins. Notably, EcTF was unable to impact the solubility of Vasp or its derivates guided proteins, suggesting its different specificities on the recognition or interaction. Additionally, this study also suggested that the translocation of Vip3 in the host cell would be regulated by the BtTF-involved model.

Improving the Named Entity Recognition of Chinese Electronic Medical Records by Combining Domain Dictionary and Rules.

Electronic medical records are an integral part of medical texts. Entity recognition of electronic medical records has triggered many studies that propose many entity extraction methods. In this paper, an entity extraction model is proposed to extract entities from Chinese Electronic Medical Records (CEMR). In the input layer of the model, we use word embedding and dictionary features embedding as input vectors, where word embedding consists of a character representation and a word representation. Then, the input vectors are fed to the bidirectional long short-term memory to capture contextual features. Finally, a conditional random field is employed to capture dependencies between neighboring tags. We performed experiments on body classification task, and the F1 values reached 90.65%. We also performed experiments on anatomic region recognition task, and the F1 values reached 93.89%. On both tasks, our model had higher performance than state-of-the-art models, such as Bi-LSTM-CRF, Bi-LSTM-Attention, and Vote. Through experiments, our model has a good effect when dealing with small frequency entities and unknown entities; with a small training dataset, our method showed 2-4% improvement on F1 value compared to the basic Bi-LSTM-CRF models. Additionally, on anatomic region recognition task, besides using our proposed entity extraction model, 12 rules we designed and domain dictionary were adopted. Then, in this task, the weighted F1 value of the three specific entities extraction reached 84.36%.

edge2vec: Representation learning using edge semantics for biomedical knowledge discovery.

BACKGROUND: Representation learning provides new and powerful graph analytical approaches and tools for the highly valued data science challenge of mining knowledge graphs. Since previous graph analytical methods have mostly focused on homogeneous graphs, an important current challenge is extending this methodology for richly heterogeneous graphs and knowledge domains. The biomedical sciences are such a domain, reflecting the complexity of biology, with entities such as genes, proteins, drugs, diseases, and phenotypes, and relationships such as gene co-expression, biochemical regulation, and biomolecular inhibition or activation. Therefore, the semantics of edges and nodes are critical for representation learning and knowledge discovery in real world biomedical problems. RESULTS: In this paper, we propose the edge2vec model, which represents graphs considering edge semantics. An edge-type transition matrix is trained by an Expectation-Maximization approach, and a stochastic gradient descent model is employed to learn node embedding on a heterogeneous graph via the trained transition matrix. edge2vec is validated on three biomedical domain tasks: biomedical entity classification, compound-gene bioactivity prediction, and biomedical information retrieval. Results show that by considering edge-types into node embedding learning in heterogeneous graphs, edge2vec significantly outperforms state-of-the-art models on all three tasks. CONCLUSIONS: We propose this method for its added value relative to existing graph analytical methodology, and in the real world context of biomedical knowledge discovery applicability.

Evaluating the granularity balance of hierarchical relationships within large biomedical terminologies towards quality improvement.

Organizing the descendants of a concept under a particular semantic relationship may be rather arbitrarily carried out during the manual creation processes of large biomedical terminologies, resulting in imbalances in relationship granularity. This work aims to propose scalable models towards systematically evaluating the granularity balance of semantic relationships. We first utilize "parallel concepts set (PCS)" and two features (the length and the strength) of the paths between PCSs to design the general evaluation models, based on which we propose eight concrete evaluation models generated by two specific types of PCSs: single concept set and symmetric concepts set. We then apply those concrete models to the IS-A relationship in FMA and SNOMED CT's Body Structure subset, as well as to the Part-Of relationship in FMA. Moreover, without loss of generality, we conduct two additional rounds of applications on the Part-Of relationship after removing length redundancies and strength redundancies sequentially. At last, we perform automatic evaluation on the imbalances detected after the final round for identifying missing concepts, misaligned relations and inconsistencies. For the IS-A relationship, 34 missing concepts, 80 misalignments and 18 redundancies in FMA as well as 28 missing concepts, 114 misalignments and 1 redundancy in SNOMED CT were uncovered. In addition, 6,801 instances of imbalances for the Part-Of relationship in FMA were also identified, including 3,246 redundancies. After removing those redundancies from FMA, the total number of Part-Of imbalances was dramatically reduced to 327, including 51 missing concepts, 294 misaligned relations, and 36 inconsistencies. Manual curation performed by the FMA project leader confirmed the effectiveness of our method in identifying curation errors. In conclusion, the granularity balance of hierarchical semantic relationship is a valuable property to check for ontology quality assurance, and the scalable evaluation models proposed in this study are effective in fulfilling this task, especially in auditing relationships with sub-hierarchies, such as the seldom evaluated Part-Of relationship.

Docetaxel-loaded nanoparticles based on star-shaped mannitol-core PLGA-TPGS diblock copolymer for breast cancer therapy.

A star-shaped biodegradable polymer, mannitol-core poly(d,l-lactide-co-glycolide)-d-α-tocopheryl polyethylene glycol 1000 succinate (M-PLGA-TPGS), was synthesized in order to provide a novel nanoformulation for breast cancer chemotherapy. This novel copolymer was prepared by a core-first approach via three stages of chemical reaction, and was characterized by nuclear magnetic resonance, gel permeation chromatography and thermogravimetric analysis. The docetaxel-loaded M-PLGA-TPGS nanoparticles (NPs), prepared by a modified nanoprecipitation method, were observed to be near-spherical shape with narrow size distribution. Confocal laser scanning microscopy showed that the uptake level of M-PLGA-TPGS NPs was higher than that of PLGA NPs and PLGA-TPGS NPs in MCF-7 cells. A significantly higher level of cytotoxicity was achieved with docetaxel-loaded M-PLGA-TPGS NPs than with commercial Taxotere®, docetaxel-loaded PLGA-TPGS and PLGA NPs. Examination of the drug loading and encapsulation efficiency proved that star-shaped M-PLGA-TPGS could carry higher levels of drug than linear polymer. The in vivo experiment showed docetaxel-loaded M-PLGA-TPGS NPs to have the highest anti-tumor efficacy. In conclusion, the star-like M-PLGA-TPGS copolymer shows potential as a promising drug-loaded biomaterial that can be applied in developing novel nanoformulations for breast cancer therapy.