pubmed2ensembl: a resource for mining the biological literature on genes

Precision Drug Repurposing: A Deep Learning Toolkit for Identifying 34 Hyperpigmentation-Associated Genes and Optimizing Treatment Selection

BACKGROUND

Hyperpigmentation is a skin disorder characterized by a localized darkening of the skin due to increased melanin production. When patients fail first line topical treatments, secondary treatments such as chemical peels and lasers are offered. However, these interventions are not devoid of risks and are associated with postinflammatory hyperpigmentation. In the quest for novel therapeutic potentials, this study aims to investigate computational methods in the identification of new targeted therapies in the treatment of hyperpigmentation.

METHODS

We used a comprehensive approach, which integrated text mining, interpreting gene lists through enrichment analysis and integration of diverse biological information (GeneCodis), protein-protein association networks and functional enrichment analyses (STRING), and plug-in network centrality parameters (Cytoscape) to pinpoint genes closely associated with hyperpigmentation. Subsequently, analysis of drug-gene interactions to identify potential drugs (Cortellis) was utilized to select drugs targeting these identified genes. Lastly, we used Deep Learning Based Drug Repurposing Toolkit (DeepPurpose) to conduct drug-target interaction predictions to ultimately identify candidate drugs with the most promising binding affinities.

RESULTS

Thirty-four hyperpigmentation-related genes were identified by text mining. Eight key genes were highlighted by utilizing GeneCodis, STRING, Cytoscape, gene enrichment, and protein-protein interaction analysis. Thirty-five drugs targeting hyperpigmentation-associated genes were identified by Cortellis, and 29 drugs, including 16 M2PK1 inhibitors, 11 KRAS inhibitors, and 2 BRAF inhibitors were recommended by DeepPurpose.

CONCLUSIONS

The study highlights the promise of advanced computational methodology for identifying potential treatments for hyperpigmentation.

Publication, funding, and experimental data in support of Human Reference Atlas construction and usage

Experts from 18 consortia are collaborating on the Human Reference Atlas (HRA) which aims to map the 37 trillion cells in the healthy human body. Information relevant for HRA construction and usage is held by experts, published in scholarly papers, and captured in experimental data. However, these data sources use different metadata schemas and cannot be cross-searched efficiently. This paper documents the compilation of a dataset, named HRAlit, that links the 136 HRA v1.4 digital objects (31 organs with 4,279 anatomical structures, 1,210 cell types, 2,089 biomarkers) to 583,117 experts; 7,103,180 publications; 896,680 funded projects, and 1,816 experimental datasets. The resulting HRAlit has 22 tables with 20,939,937 records including 6 junction tables with 13,170,651 relationships. The HRAlit can be mined to identify leading experts, major papers, funding trends, or alignment with existing ontologies in support of systematic HRA construction and usage.

Identification of Potential Therapeutic Targets for Myopic Choroidal Neovascularization via Discovery-Driven Data Mining

Purpose: Myopic choroidal neovascularization (mCNV) is a prevalent cause of vision loss. However, the development of effective therapeutic targets for mCNV has been hindered by the paucity of suitable animal models. Therefore, the aim of this study is to identify potential genes and pathways associated with mCNV and to unearth prospective therapeutic targets that can be utilized to devise efficacious treatments.Methods: Text data mining was used to identify genes linked to choroid, neovascularization, and myopia. g: Profiler was utilized to analyze the biological processes of gene ontology and the Reactome pathways. Protein interaction network analysis was performed using strings and visualized in Cytoscape. MCODE and cytoHubba were used for further screening.Results: Discovery-driven text data mining identified 55 potential genes related to choroid, neovascularization, and myopia. Gene enrichment analysis revealed 11 biological processes and seven Reactome pathways. A protein-protein interaction network with 47 nodes was constructed and analyzed using centrality ranking. Key clusters were identified through algorithm tools. Finally, 14 genes (IL6, FGF2, MMP9, IL10, TNF, MMP2, HGF, MMP3, IGF1, CCL2, CTNNB1, BDNF, NGF, and EDN1), in addition to VEGFA, were evaluated as targets with potential as future therapeutics.Conclusions: This study provides new potential therapeutic targets for mCNV, including IL6, FGF2, MMP9, IL10, TNF, MMP2, HGF, MMP3, IGF1, CCL2, CTNNB1, BDNF, NGF, and EDN1, which correspond to seven potential enriched pathways. These findings provide a basis for further research and offer new possibilities for developing therapeutic interventions for this condition.

Investigation of anti-depression effects and potential mechanisms of the ethyl acetate extract of Cynomorium songaricum Rupr. through the integration of in vivo experiments, LC-MS/MS chemical analysis, and a systems biology approach

Background: Cynomorium songaricum Rupr. has long been used as an anti-inflammatory, antidepressant, and anti-aging agent in traditional Chinese medicine in Asia. Its ethyl acetate extract (ECS) has been identified as the main antioxidant component with neuroprotective and estrogen-like effects. However, the potential of ECS in treating depression has not been explored yet. Methods: We identified the primary metabolites in ECS in this study using liquid chromatography-electrospray tandem mass spectrometry (LC-MS/MS). Network analysis was used to find the potential targets and pathways associated with the anti-neuroinflammatory depression action of the ECS. In addition, we established a corticosterone (CORT)-induced depression mouse model to assess ECS's antidepressant effects by monitoring various behavioral changes (e.g., sucrose preference, forced swimming, tail suspension, and open field tests) and biochemical indices of the hippocampus, and validating the network analysis results. Significant pathways underwent verification through western blotting based on network analysis prediction. Results: Our study demonstrates that ECS possesses significant antidepressant activity. The LC-MS/MS analysis of ECS identified 30 main metabolites, including phloridzin, phlorizin, ursolic acid, and naringenin, as well as other flavonoids, terpenoids, and phenolic acids. These metabolites were found to be associated with 64 candidate target proteins related to neuroinflammatory depression from the database, and ten hub proteins were identified through filtration: CXCL8, ICAM1, NOS2, SELP, TNF, IL6, APP, ACHE, MAOA and ADA. Functional enrichment analyses of the candidate targets revealed their primary roles in regulating cytokine production, inflammatory response, cytokine activity, and tumor necrosis factor receptor binding. In vivo, ECS improved hippocampal neuroinflammation in the mouse model. Specifically, ECS reduced the expression of inflammatory factors in the hippocampus, inhibited M1 microglial cell polarization, and alleviated depression through the regulation of the NF-κB-NLRP3 inflammation pathway. Conclusion: Based on experimental and network analysis, this study revealed for the first time that ECS exerted antidepression effect via anti-neuroinflammation. Our research provides valuable information on the use of ECS as an alternative therapeutic approach for depression.

Molecular Mechanisms of Resistance to Ionizing Radiation in S. cerevisiae and Its Relationship with Aging, Oxidative Stress, and Antioxidant Activity

The repair of the damage produced to the genome and proteome by the action of ionizing radiation, oxidizing agents, and during aging is important to maintain cellular homeostasis. Many of the metabolic pathways influence multiple processes. In this way, this work aims to study the relationship between resistance/response to ionizing radiation, cellular aging, and the response mechanisms to oxidative stress, free radicals, reactive oxygen species (ROS), and antioxidant activity in the yeast S. cerevisiae. Systems biology allows us to use tools that reveal the molecular mechanisms common to different cellular response phenomena. The results found indicate that homologous recombination, non-homologous end joining, and base excision repair pathways are the most important common processes necessary to maintain cellular homeostasis. The metabolic routes of longevity regulation are those that jointly contribute to the three phenomena studied. This study proposes eleven common biomarkers for response/resistance to ionizing radiation and aging (EXO1, MEC1, MRE11, RAD27, RAD50, RAD51, RAD52, RAD55, RAD9, SGS1, YKU70) and two biomarkers for response/resistance to radiation and oxidative stress, free radicals, ROS, and antioxidant activity (NTG1, OGG1). In addition, it is important to highlight that the HSP104 protein could be a good biomarker common to the three phenomena studied.

Potential alternative drug treatment for bone giant cell tumor

Background: Bone giant cell tumor (BGCT) is one of the world's major disease types of locally aggressive bone tumors. In recent years, denosumab treatment has been introduced before curettage surgery. However, the current therapeutic was practical only sometimes, given the local recurrence effects after discontinuation of denosumab. Due to the complex nature of BGCT, this study aims to use bioinformatics to identify potential genes and drugs associated with BGCT. Methods: The genes that integrate BGCT and fracture healing were determined by text mining. The gene was obtained from the pubmed2ensembl website. We filtered out common genes for the function, and signal pathway enrichment analyses were implemented. The protein-protein interaction (PPI) networks and the hub genes were screened by MCODE built-in Cytoscape software. Lastly, the confirmed genes were queried in the Drug Gene Interaction Database to determine potential genes and drugs. Results: Our study finally identified 123 common specific genes in bone giant cell tumors and fracture healing text mining concepts. The GO enrichment analysis finally analyzed 115 characteristic genes in BP, CC, and MF. We selected 10 KEGG pathways and identified 68 characteristic genes. We performed protein-protein interaction analysis (PPI) on 68 selected genes and finally identified seven central genes. In this study, these seven genes were substituted into drug-gene interactions, and there were 15 antineoplastic drugs, 1 anti-involving drug, and 1 anti-influenza drug. Conclusion: The 7 genes (including ANGPT2, COL1A1, COL1A2, CTSK, FGFR1, NTRK2, and PDGFB) and 17 drugs, which have not been used in BGCT, but 6 of them approved by the FDA for other diseases, could be potential genes and drugs, respectively, to improve BGCT treatment. In addition, the correlation study and analysis of potential drugs through genes provide great opportunities to promote the repositioning of drugs and the study of pharmacology in the pharmaceutical industry.

Drug Discovery in Canine Pyometra Disease Identified by Text Mining and Microarray Data Analysis

Canine pyometra, which is accompanied by bacterial contamination of the dog uterus, is defined as a complex disease associated with the activation of several systems, including the immune system. This study uses text mining and microarray data analysis methods to discover some existing targeted gene drugs and expand potential new drug indications. Text mining ("canine pyometra") and microarray data analysis (GSE99877) were used to obtain a common set of genes. These genes and protein-protein interaction (PPI) networks were analyzed using Gene Ontology and the Kyoto Encyclopedia of Genes and Genomes. Then, the important genes clustered in the PPI network were selected for gene-drug interaction analysis to provide evidence for potential drug discovery. Through text mining and data analysis, we obtained 17,544 text mining genes (TMGs) and 399 differentially expressed genes (DEGs), respectively. There were 256 repeat genes between TMGs and DEGs, including 70 upregulated genes and 186 downregulated genes. Thirty-seven genes clustered in three significant gene modules. Eight of the 37 genes can target 23 existing drugs. In conclusion, the discovery of 8 immune response-related genes (BTK, CSF2RA, CSF2RB, ITGAL, NCF4, PLCG2, PTPRC, and TOP2A) targeting 23 existing drugs may expand the drug indications for pyometra-related diseases in dogs.

A Four-Gene Signature Associated with Radioresistance in Head and Neck Squamous Cell Carcinoma Identified by Text Mining and Data Analysis

Purpose

Head and neck squamous cell carcinoma (HNSCC) is the sixth leading cancer globally, and radiotherapy plays a crucial part in its treatment. This study was designed to identify potential genes related to radiation resistance in HNSCC.

Method

We first used text mining to obtain common genes related to radiotherapy resistance and HNSCC in published articles. Functional enrichment analyses were conducted to identify the significantly enriched pathways and genes. Protein and protein interactions were performed, and the most significant gene modules were determined; then, genes in the gene modules were validated at transcriptional levels and overall survival. Gene set variation analysis (GSVA) score was calculated, and the association between GSVA score and survival/pathway was estimated. Immune cell infiltration, methylation, and genetic alteration analysis of these genes was conducted in HNSCC patients. Finally, potential sensitive anticancer drugs related to target genes were obtained.

Result

We identified 583 common genes through text mining. After further validation, a four-gene signature (EPHB2, SPP1, SERPINE1, and VEGFC) was constructed. The patients with higher GSVA scores have a worse prognosis than those with lower GSVA scores. Differences in methylation of these four genes in HNSCC tumor tissue and normal tissue were compared, with higher methylation levels of EBPH2 and SPP1 in normal tissue and higher methylation levels of SERPINE1 in the tumor. Immune cell infiltration revealed that the increased expression of these genes was closely related to the infiltration level of CD4+ T cell, neutrophil, macrophage, and dendritic cell. Thirty drugs, including 22 positively and eight negatively correlated drugs that most correlated with related genes, were available for treating HNSCC.

Conclusion

In this study, we identified four potential genes as well as corresponding drugs that might be related to radioresistance in HNSCC patients. These candidate genes may provide a promising avenue to further elevate radiotherapy efficacy.

Text Mining-Based Drug Discovery for Connective Tissue Disease–Associated Pulmonary Arterial Hypertension

Background: The current medical treatments for connective tissue disease–associated pulmonary arterial hypertension (CTD-PAH) do not show favorable efficiency for all patients, and identification of novel drugs is desired.

Methods: Text mining was performed to obtain CTD- and PAH-related gene sets, and the intersection of the two gene sets was analyzed for functional enrichment through DAVID. The protein–protein interaction network of the overlapping genes and the significant gene modules were determined using STRING. The enriched candidate genes were further analyzed by Drug Gene Interaction database to identify drugs with potential therapeutic effects on CTD-PAH.

Results: Based on text mining analysis, 179 genes related to CTD and PAH were identified. Through enrichment analysis of the genes, 20 genes representing six pathways were obtained. To further narrow the scope of potential existing drugs, we selected targeted drugs with a Query Score ≥5 and Interaction Score ≥1. Finally, 13 drugs targeting the six genes were selected as candidate drugs, which were divided into four drug–gene interaction types, and 12 of them had initial drug indications approved by the FDA. The potential gene targets of the drugs on this list are IL-6 (one drug) and IL-1β (two drugs), MMP9 (one drug), VEGFA (three drugs), TGFB1 (one drug), and EGFR (five drugs). These drugs might be used to treat CTD-PAH.

Conclusion: We identified 13 drugs targeting six genes that may have potential therapeutic effects on CTD-PAH.

A Computational Text Mining-Guided Meta-Analysis Approach to Identify Potential Xerostomia Drug Targets

Xerostomia (subjective complaint of dry mouth) is commonly associated with salivary gland hypofunction. Molecular mechanisms associated with xerostomia pathobiology are poorly understood, thus hampering drug development. Our objectives were to (i) use text-mining tools to investigate xerostomia and dry mouth concepts, (ii) identify associated molecular interactions involving genes as candidate drug targets, and (iii) determine how drugs currently used in clinical trials may impact these genes and associated pathways. PubMed and PubMed Central were used to identify search terms associated with xerostomia and/or dry mouth. Search terms were queried in pubmed2ensembl. Protein-protein interaction (PPI) networks were determined using the gene/protein network visualization program search tool for recurring instances of neighboring genes (STRING). A similar program, Cytoscape, was used to determine PPIs of overlapping gene sets. The drug-gene interaction database (DGIdb) and the clinicaltrials.gov database were used to identify potential drug targets from the xerostomia/dry mouth PPI gene set. We identified 64 search terms in common between xerostomia and dry mouth. STRING confirmed PPIs between identified genes (CL = 0.90). Cytoscape analysis determined 58 shared genes, with cytokine-cytokine receptor interaction representing the most significant pathway (p = 1.29 × 10-23) found in the Kyoto encyclopedia of genes and genomes (KEGG). Fifty-four genes in common had drug interactions, per DGIdb analysis. Eighteen drugs, targeting the xerostomia/dry mouth PPI network, have been evaluated for xerostomia, head and neck cancer oral complications, and Sjögren's Syndrome. The PPI network genes IL6R, EGFR, NFKB1, MPO, and TNFSF13B constitute a possible biomarker signature of xerostomia. Validation of the candidate biomarkers is necessary to better stratify patients at the genetic and molecular levels to facilitate drug development or to monitor response to treatment.

A Hybrid Protocol for Finding Novel Gene Targets for Various Diseases Using Microarray Expression Data Analysis and Text Mining

The advancement in technology for various scientific experiments and the amount of raw data produced from that is enormous, thus giving rise to various subsets of biologists working with genome, proteome, transcriptome, expression, pathway, and so on. This has led to exponential growth in scientific literature which is becoming beyond the means of manual curation and annotation for extracting information of importance. Microarray data are expression data, analysis of which results in a set of up/downregulated lists of genes that are functionally annotated to ascertain the biological meaning of genes. These genes are represented as vocabularies and/or Gene Ontology terms when associated with pathway enrichment analysis need relational and conceptual understanding to a disease. The chapter deals with a hybrid approach we designed for identifying novel drug-disease targets. Microarray data for muscular dystrophy is explored here as an example and text mining approaches are utilized with an aim to identify promisingly novel drug targets. Our main objective is to give a basic overview from a biologist's perspective for whom text mining approaches of data mining and information retrieval is fairly a new concept. The chapter aims to bridge the gap between biologist and computational text miners and bring about unison for a more informative research in a fast and time efficient manner.

Searching for essential genes and drug discovery in breast cancer and periodontitis via text mining and bioinformatics analysis

The primary purpose of the study was (1) to search for the essential genes associated with breast cancer and periodontitis, and (2) to identify candidate drugs targeted to these genes for expanding the potential drug indications. The genes related to both breast cancer and periodontitis were determined by text mining. Gene ontology and Kyoto Encyclopedia of Genes and Genomes analysis were performed on these genes, and protein-protein interaction analysis was carried out to export significant module genes. Drug-gene interaction database was employed for potential drug discovery. We identified 221 genes common to both breast cancer and periodontitis. The top six significant enrichment terms and 15 enriched signal pathways were selected. Among 24 significant genes demonstrated as a gene cluster, we found SERPINA1 and TF were significantly related to poor overall survival between the relatively high and low groups in patients. Using the final two genes, 12 drugs were identified that had potential therapeutic effects. SERPINA1 and TF were screened out as essential genes related to both breast cancer and periodontitis, targeting 12 candidate drugs that may expand drug indications. Drug discovery using text mining and analysis of different databases can promote the identification of existing drugs that have the potential of administration to improve treatment in breast cancer.

Association between chronic periodontitis and the risk of Alzheimer's disease: combination of text mining and GEO dataset

BACKGROUND

Although chronic periodontitis has previously been reported to be linked with Alzheimer's disease (AD), the pathogenesis between the two is unclear. The purpose of this study is to analyze and screen the relevant and promising molecular markers between chronic periodontitis and Alzheimer's disease (AD).

METHODS

In this paper, we analyzed three AD expression datasets and extracted differentially expressed genes (DEGs), then intersected them with chronic periodontitis genes obtained from text mining, and finally obtained integrated DEGs. We followed that by enriching the matching the matching cell signal cascade through DAVID analysis. Moreover, the MCODE of Cytoscape software was employed to uncover the protein-protein interaction (PPI) network and the matching hub gene. Finally, we verified our data using a different independent AD cohort.

RESULTS

The chronic periodontitis gene set acquired from text abstracting was intersected with the previously obtained three AD groups, and 12 common genes were obtained. Functional enrichment assessment uncovered 12 cross-genes, which were mainly linked to cell morphogenesis involved in neuron differentiation, leading edge membrane, and receptor ligand activity. After PPI network creation, the ten hub genes linked to AD were retrieved, consisting of SPP1, THY1, CD44, ITGB1, HSPB3, CREB1, SST, UCHL1, CCL5 and BMP7. Finally, the function terms in the new independent dataset were used to verify the previous dataset, and we found 22 GO terms and one pathway, "ECM-receptor interaction pathways", in the overlapping functional terms.

CONCLUSIONS

The establishment of the above-mentioned candidate key genes, as well as the enriched signaling cascades, provides promising molecular markers for chronic periodontitis-related AD, which may help the diagnosis and treatment of AD patients in the future.

A Computational Exploration of the Molecular Network Associated to Neuroinflammation in Alzheimer's Disease

Neuroinflammation, as defined by the presence of classically activated microglia, is thought to play a key role in numerous neurodegenerative disorders such as Alzheimer’s disease. While modulating neuroinflammation could prove beneficial against neurodegeneration, identifying its most relevant biological processes and pharmacological targets remains highly challenging. In the present study, we combined text-mining, functional enrichment and protein-level functional interaction analyses to 1) identify the proteins significantly associated to neuroinflammation in Alzheimer’s disease over the scientific literature, 2) distinguish the key proteins most likely to control the neuroinflammatory processes significantly associated to Alzheimer's disease, 3) identify their regulatory microRNAs among those dysregulated in Alzheimer's disease and 4) assess their pharmacological targetability. 94 proteins were found to be significantly associated to neuroinflammation in Alzheimer’s disease over the scientific literature and IL4, IL10 and IL13 signaling as well as TLR-mediated MyD88- and TRAF6-dependent responses were their most significantly enriched biological processes. IL10, TLR4, IL6, AKT1, CRP, IL4, CXCL8, TNF-alpha, ITGAM, CCL2 and NOS3 were identified as the most potent regulators of the functional interaction network formed by these immune processes. These key proteins were indexed to be regulated by 63 microRNAs dysregulated in Alzheimer's disease, 13 long non-coding RNAs and targetable by 55 small molecules and 8 protein-based therapeutics. In conclusion, our study identifies eleven key proteins with the highest ability to control neuroinflammatory processes significantly associated to Alzheimer’s disease, as well as pharmacological compounds with single or pleiotropic actions acting on them. As such, it may facilitate the prioritization of diagnostic and target-engagement biomarkers as well as the development of effective therapeutic strategies against neuroinflammation in Alzheimer’s disease.

Extracting knowledge networks from plant scientific literature: potato tuber flesh color as an exemplary trait

BACKGROUND

Scientific literature carries a wealth of information crucial for research, but only a fraction of it is present as structured information in databases and therefore can be analyzed using traditional data analysis tools. Natural language processing (NLP) is often and successfully employed to support humans by distilling relevant information from large corpora of free text and structuring it in a way that lends itself to further computational analyses. For this pilot, we developed a pipeline that uses NLP on biological literature to produce knowledge networks. We focused on the flesh color of potato, a well-studied trait with known associations, and we investigated whether these knowledge networks can assist us in formulating new hypotheses on the underlying biological processes.

RESULTS

We trained an NLP model based on a manually annotated corpus of 34 full-text potato articles, to recognize relevant biological entities and relationships between them in text (genes, proteins, metabolites and traits). This model detected the number of biological entities with a precision of 97.65% and a recall of 88.91% on the training set. We conducted a time series analysis on 4023 PubMed abstract of plant genetics-based articles which focus on 4 major Solanaceous crops (tomato, potato, eggplant and capsicum), to determine that the networks contained both previously known and contemporaneously unknown leads to subsequently discovered biological phenomena relating to flesh color. A novel time-based analysis of these networks indicates a connection between our trait and a candidate gene (zeaxanthin epoxidase) already two years prior to explicit statements of that connection in the literature.

CONCLUSIONS

Our time-based analysis indicates that network-assisted hypothesis generation shows promise for knowledge discovery, data integration and hypothesis generation in scientific research.

Text Mining-Based Drug Discovery in Osteoarthritis

Background

Osteoarthritis (OA) is a chronic and degenerative joint disease, which causes stiffness, pain, and decreased function. At the early stage of OA, nonsteroidal anti-inflammatory drugs (NSAIDs) are considered the first-line treatment. However, the efficacy and utility of available drug therapies are limited. We aim to use bioinformatics to identify potential genes and drugs associated with OA.

Methods

The genes related to OA and NSAIDs therapy were determined by text mining. Then, the common genes were performed for GO, KEGG pathway analysis, and protein-protein interaction (PPI) network analysis. Using the MCODE plugin-obtained hub genes, the expression levels of hub genes were verified using quantitative real-time polymerase chain reaction (qRT-PCR). The confirmed genes were queried in the Drug Gene Interaction Database to determine potential genes and drugs.

Results

The qRT-PCR result showed that the expression level of 15 genes was significantly increased in OA samples. Finally, eight potential genes were targetable to a total of 53 drugs, twenty-one of which have been employed to treat OA and 32 drugs have not yet been used in OA.

Conclusions

The 15 genes (including PTGS2, NLRP3, MMP9, IL1RN, CCL2, TNF, IL10, CD40, IL6, NGF, TP53, RELA, BCL2L1, VEGFA, and NOTCH1) and 32 drugs, which have not been used in OA but approved by the FDA for other diseases, could be potential genes and drugs, respectively, to improve OA treatment. Additionally, those methods provided tremendous opportunities to facilitate drug repositioning efforts and study novel target pharmacology in the pharmaceutical industry.

Identification of drug compounds for keloids and hypertrophic scars: drug discovery based on text mining and DeepPurpose

Background

Keloids (KL) and hypertrophic scars (HS) are forms of abnormal cutaneous scarring characterized by excessive deposition of extracellular matrix and fibroblast proliferation. Currently, the efficacy of drug therapies for KL and HS is limited. The present study aimed to investigate new drug therapies for KL and HS by using computational methods.

Methods

Text mining and GeneCodis were used to mine genes closely related to KL and HS. Protein-protein interaction analysis was performed using Search Tool for the Retrieval of Interacting Genes/Proteins (STRING) and Cytoscape. The selection of drugs targeting the genes closely related to KL and HS was carried out using Pharmaprojects. Drug-target interaction prediction was performed using DeepPurpose, through which candidate drugs with the highest predicted binding affinity were finally obtained.

Results

Our analysis using text mining identified 69 KL- and HS-related genes. Gene enrichment analysis generated 25 genes, representing 7 pathways and 130 targeting drugs. DeepPurpose recommended 14 drugs as the final drug list, including 2 phosphatidylinositol-4,5-bisphosphate 3-kinase (PI3K) inhibitors, 10 prostaglandin-endoperoxide synthase 2 (PTGS2) inhibitors and 2 vascular endothelial growth factor A (VEGFA) antagonists.

Conclusions

Drug discovery using in silico text mining and DeepPurpose may be a powerful and effective way to identify drugs targeting the genes related to KL and HS.

Kindlin-2 regulates the differentiation of 3T3-L1 preadipocytes: implications for wound healing

Background

Adipose tissue has been proven to play a crucial role in wound healing, while kindlin-2, an integrin-associated protein, has been shown to regulate cell adhesion, migration, and differentiation. This study aimed to explore its involvement in the cell differentiation of 3T3-L1 preadipocytes and its role in wound healing.

Methods

Cell adhesion, Cell Counting Kit-8 (CCK-8), Transwell, and in vitro wound healing assays, along with adipogenic and osteogenic differentiation induction were performed in 3T3-L1 preadipocytes in which kindlin-2 was knocked down or overexpressed. In vivo, kindlin-2 (+/−) transgenic mice were constructed, and wound healing was analyzed by immunohistochemistry (IHC) in a mouse dorsal wound model. Real-time polymerase chain reaction (RT-PCR) and western blotting were performed to analyze the expression of adipokines and adipogenic markers in mouse wound tissues. Adipogenic differentiation induction of adipose tissue stromal vascular fraction (SVF) were performed, and the expression of adipogenic markers in SVF was detected by western blotting. The target signaling pathway highly related to adipogenic differentiation was explored by computational biology and verified by western blotting.

Results

Knockdown of kindlin-2 was found to inhibit the adhesion, migration, and adipogenic differentiation of 3T3-L1 preadipocytes while promoting their osteogenic differentiation. In contrast, kindlin-2 overexpression resulted in increased adhesion, migration, and adipogenic differentiation of 3T3-L1 preadipocytes while reducing osteogenic differentiation. In vivo, downregulation of kindlin-2 inhibited adipogenesis in kindlin-2 transgenic mice, resulting in delayed wound healing by inhibiting inflammation, angiogenesis, collagen remodeling, and wound contraction. Mechanistically, we found that kindlin-2 could regulate adipogenic differentiation through PI3K/AKT/mTOR signaling pathway.

Conclusions

Our study revealed the essential role that kindlin-2 has in the differentiation and wound healing of 3T3-L1 preadipocytes, which offers a theoretical basis for further research and a novel strategy for wound healing.

Identification of key pathways and differentially expressed genes in bronchopulmonary dysplasia using bioinformatics analysis

OBJECTIVES

The objective of this study was to discover unknown differentially expressed genes (DEGs) associated with bronchopulmonary dysplasia (BPD), analyze their functions and enriched signaling pathways, and identify hub genes correlating with BPD incidence and evolvement.

RESULTS

Of 1289 DEGs identified, 568 were downregulated and 721 were upregulated. The DEGs were mainly associated with oxidative stress, angiogenesis, extracellular matrix, inflammation, cell cycle, and protein binding. Eight DEGs were identified as hub genes, including C-X-C motif chemokine ligand 5 (Cxcl5), connective tissue growth factor (Ctgf), interleukin 6 (IL6), matrix metallopeptidase 9 (Mmp9), mitogen-activated protein kinase 14 (Mapk14), platelet and endothelial cell adhesion molecule 1 (Pecam1), TIMP metallopeptidase inhibitor 1 (Timp1), and TIMP metallopeptidase inhibitor 2 (Timp2). IL6 mRNA and protein expression levels were significantly increased in the peripheral blood of neonates with BPD.

CONCLUSIONS

Hence, BPD involves complex biological changes. Our findings indicate that inflammation and angiogenesis may play major roles in BPD pathogenesis and that IL6 has the potential to serve as a biomarker for early BPD diagnosis.

Integrative analysis of key candidate genes and signaling pathways in autoimmune thyroid dysfunction related to anti-CTLA-4 therapy by bioinformatics

Cytotoxic T lymphocyte-associated antigen-4 (CTLA-4), the first immune checkpoint to be targeted clinically, has provided an effective treatment option for various malignancies. However, the clinical advantages associated with CTLA-4 inhibitors can be offset by the potentially severe immune-related adverse events (IRAEs), including autoimmune thyroid dysfunction. To investigate the candidate genes and signaling pathways involving in autoimmune thyroid dysfunction related to anti-CTLA-4 therapy, integrated differentially expressed genes (DEGs) were extracted from the intersection of genes from Gene Expression Omnibus (GEO) datasets and text mining. The functional enrichment was performed by gene ontology (GO) annotation and Kyoto encyclopedia of genes and genomes (KEGG) pathway analysis. Protein-protein interaction (PPI) network, module enrichment, and hub gene identification were constructed and visualized by the online Search Tool for the Retrieval of Interacting Genes (STRING) and Cytoscape software. A total of 22 and 17 integrated human DEGs in hypothyroidism and hyperthyroidism group related to anti-CTLA-4 therapy were identified, respectively. Functional enrichment analysis revealed 24 GO terms and 1 KEGG pathways in the hypothyroid group and 21 GO terms and 2 KEGG pathways in the hyperthyroid group. After PPI network construction, the top five hub genes associated with hypothyroidism were extracted, including ALB, MAPK1, SPP1, PPARG, and MIF, whereas those associated with hyperthyroidism were ALB, FCGR2B, CD44, LCN2, and CD74. The identification of the candidate key genes and enriched signaling pathways provides potential biomarkers for autoimmune thyroid dysfunction related to anti-CTLA-4 therapy and might contribute to the future diagnosis and management of IRAEs for cancer patients.

In silico analysis of the molecular regulatory networks in peripheral arterial occlusive disease

BACKGROUND

Peripheral arterial occlusive disease (PAOD) is a global public health concern that decreases the quality of life of the patients and can lead to disabilities and death. The aim of this study was to identify the genes and pathways associated with PAOD pathogenesis, and the potential therapeutic targets.

METHODS

Differentially expressed genes (DEGs) and miRNAs related to PAOD were extracted from the GSE57691 dataset and through text mining. Additionally, bioinformatics analysis was applied to explore gene ontology, pathways and protein-protein interaction of those DEGs. The potential miRNAs targeting the DEGs and the transcription factors (TFs) regulating miRNAs were predicted by multiple different databases.

RESULTS

A total of 59 DEGs were identified, which were significantly enriched in the inflammatory response, immune response, chemokine-mediated signaling pathway and JAK-STAT signaling pathway. Thirteen genes including IL6, CXCL12, IL1B, and STAT3 were hub genes in protein-protein interaction network. In addition, 513 miRNA-target gene pairs were identified, of which CXCL12 and PTPN11 were the potential targets of miRNA-143, and IL1B of miRNA-21. STAT3 was differentially expressed and regulated 27 potential target miRNAs including miRNA-143 and miRNA-21 in TF-miRNA regulatory network.

CONCLUSION

In summary, inflammation, immune response and STAT3-mediated miRNA-target genes axis play an important role in PAOD development and progression.

Hydroxychloroquine suppresses lung tumorigenesis via inducing FoxO3a nuclear translocation through STAT3 inactivation

BACKGROUND

Hydroxychloroquine exhibits synergistic anticancer properties as an adjuvant. However, the role and molecular mechanisms underlying of HCQ as monotherapy for lung adenocarcinoma (LUAD) have yet to be elucidated.

METHODS

We assessed the antitumor effects of HCQ in LUAD cells through a series of in vitro and in vivo assays. GEO database and R packages were used to predict molecular mechanisms of HCQ in the treatment of lung adenocarcinoma, followed by verification of gene expression and subcellular localization via immunoblotting, immunofluorescent and immunohistochemistry assays.

RESULTS

We showed the phenotypic effects that HCQ inhibited cell growth, induced apoptosis and cell cycle arrest at G1/S transition in A549 and PC-9 cells, which was associated with inhibition of CDK2, CDK4, CyclinD1 and CyclinE, but up-regulation of p21 and p27^Kip1. Bioinformatic analysis predicted that 63 targets related to HCQ and LUAD were mainly enriched in JAK-STAT and FoxO pathways. Then, we observed that HCQ decreased the phosphorylation of STAT3, but increased the expression of FoxO3a and its accumulation in the nucleus. The specific STAT3 inhibitor cryptotanshinon augmented the HCQ-induced upregulation and nuclear translocation of FoxO3a. In addition, HCQ increased the expression of p27^Kip1, which was impaired by FoxO3a blockade with siRNA. Finally, ablation of p27^Kip1 expression abrogated the cytotoxicity of HCQ. More importantly, similar results were further confirmed in vivo.

CONCLUSIONS

Taken together, this study suggests that STAT3/FoxO3a/p27^Kip1 signaling pathway is involved in the anticancer effects of HCQ, and provides preliminary evidence for therapeutic prospects of HCQ alone in LUAD.

Identification of key candidate genes and molecular pathways in white fat browning: an anti-obesity drug discovery based on computational biology

Background

Obesity—with its increased risk of obesity-associated metabolic diseases—has become one of the greatest public health epidemics of the twenty-first century in affluent countries. To date, there are no ideal drugs for treating obesity. Studies have shown that activation of brown adipose tissue (BAT) can promote energy consumption and inhibit obesity, which makes browning of white adipose tissue (WAT) a potential therapeutic target for obesity. Our objective was to identify genes and molecular pathways associated with WAT and the activation of BAT to WAT browning, by using publicly available data and computational tools; this knowledge might help in targeting relevant signaling pathways for treating obesity and other related metabolic diseases.

Results

In this study, we used text mining to find out genes related to brown fat and white fat browning. Combined with biological process and pathway analysis in GeneCodis and protein-protein interaction analysis by using STRING and Cytoscape, a list of high priority target genes was developed. The Human Protein Atlas was used to analyze protein expression. Candidate drugs were derived on the basis of the drug-gene interaction analysis of the final genes. Our study identified 18 genes representing 6 different pathways, targetable by a total of 33 drugs as possible drug treatments. The final list included 18 peroxisome proliferator-activated receptor gamma (PPAR-γ) agonists, 4 beta 3 adrenoceptor (β3-AR) agonists, 1 insulin sensitizer, 3 insulins, 6 lipase clearing factor stimulants and other drugs.

Conclusions

Drug discovery using in silico text mining, pathway, and protein-protein interaction analysis tools may be a method of exploring drugs targeting the activation of brown fat or white fat browning, which provides a basis for the development of novel targeted therapies as potential treatments for obesity and related metabolic diseases.

Identification of differentially expressed genes and signaling pathways using bioinformatics in interstitial lung disease due to tyrosine kinase inhibitors targeting the epidermal growth factor receptor

Interstitial lung disease (ILD) is a rare but lethal adverse effect of epidermal growth factor receptor (EGFR) tyrosine kinase inhibitors (TKIs) treatment. The specific mechanism of this disease is not fully understood. To systematically analyze genes associated with EGFR-TKI induced ILD, gene data of EGFR-TKI induced ILD were extracted initially using text mining, and then the intersection between genes from text mining and Gene Expression Omnibus (GEO) dataset was taken for further protein-protein interaction (PPI) analysis using String-bd database. Go ontology (GO) and pathway enrichment analysis was also conducted based on Database of Annotation, Visualization and Integrated Discovery (DAVID) platform. The PPI network generated by STRING was visualized by Cytoscape, and the topology scores, functional regions and gene annotations were analyzed using plugins of CytoNCA, molecular complex detection (MCODE) and ClueGo. 37 genes were identified as EGFR-TKI induced ILD related. Gene enrichment analysis yield 18 enriched GO terms and 12 associated pathways. A PPI network that included 199 interactions for a total of 35 genes was constructed. Ten genes were selected as hub genes using CytoNCA plugin, and four highly connected clusters were identified using MCODE plugin. GO and pathway annotation analysis for the cluster one revealed that five genes were associated with either response to dexamethasone or with lung fibrosis, including CTGF, CCL2, IGF1, EGFR and ICAM1. Our data might be useful to reveal the pathological mechanisms of EGFR-TKI induced ILD and provide evidence for the diagnosis and treatment in the future.

Text mining-based in silico drug discovery in oral mucositis caused by high-dose cancer therapy

INTRODUCTION

Oral mucositis (OM) is a major dose-limiting side effect of chemotherapy and radiation used in cancer treatment. Due to the complex nature of OM, currently available drug-based treatments are of limited efficacy.

OBJECTIVES

Our objectives were (i) to determine genes and molecular pathways associated with OM and wound healing using computational tools and publicly available data and (ii) to identify drugs formulated for topical use targeting the relevant OM molecular pathways.

METHODS

OM and wound healing-associated genes were determined by text mining, and the intersection of the two gene sets was selected for gene ontology analysis using the GeneCodis program. Protein interaction network analysis was performed using STRING-db. Enriched gene sets belonging to the identified pathways were queried against the Drug-Gene Interaction database to find drug candidates for topical use in OM.

RESULTS

Our analysis identified 447 genes common to both the "OM" and "wound healing" text mining concepts. Gene enrichment analysis yielded 20 genes representing six pathways and targetable by a total of 32 drugs which could possibly be formulated for topical application. A manual search on ClinicalTrials.gov confirmed no relevant pathway/drug candidate had been overlooked. Twenty-five of the 32 drugs can directly affect the PTGS2 (COX-2) pathway, the pathway that has been targeted in previous clinical trials with limited success.

CONCLUSIONS

Drug discovery using in silico text mining and pathway analysis tools can facilitate the identification of existing drugs that have the potential of topical administration to improve OM treatment.

Applying Monte Carlo Simulation to Biomedical Literature to Approximate Genetic Network

Biologists often need to know the set of genes associated with a given set of genes or a given disease. We propose in this paper a classifier system called Monte Carlo for Genetic Network (MCforGN) that can construct genetic networks, identify functionally related genes, and predict gene-disease associations. MCforGN identifies functionally related genes based on their co-occurrences in the abstracts of biomedical literature. For a given gene g , the system first extracts the set of genes found within the abstracts of biomedical literature associated with g. It then ranks these genes to determine the ones with high co-occurrences with g . It overcomes the limitations of current approaches that employ analytical deterministic algorithms by applying Monte Carlo Simulation to approximate genetic networks. It does so by conducting repeated random sampling to obtain numerical results and to optimize these results. Moreover, it analyzes results to obtain the probabilities of different genes' co-occurrences using series of statistical tests. MCforGN can detect gene-disease associations by employing a combination of centrality measures (to identify the central genes in disease-specific genetic networks) and Monte Carlo Simulation. MCforGN aims at enhancing state-of-the-art biological text mining by applying novel extraction techniques. We evaluated MCforGN by comparing it experimentally with nine approaches. Results showed marked improvement.

Contribution of bioinformatics prediction in microRNA-based cancer therapeutics

Despite enormous efforts, cancer remains one of the most lethal diseases in the world. With the advancement of high throughput technologies massive amounts of cancer data can be accessed and analyzed. Bioinformatics provides a platform to assist biologists in developing minimally invasive biomarkers to detect cancer, and in designing effective personalized therapies to treat cancer patients. Still, the early diagnosis, prognosis, and treatment of cancer are an open challenge for the research community. MicroRNAs (miRNAs) are small non-coding RNAs that serve to regulate gene expression. The discovery of deregulated miRNAs in cancer cells and tissues has led many to investigate the use of miRNAs as potential biomarkers for early detection, and as a therapeutic agent to treat cancer. Here we describe advancements in computational approaches to predict miRNAs and their targets, and discuss the role of bioinformatics in studying miRNAs in the context of human cancer.

Mining locus tags in PubMed Central to improve microbial gene annotation

BACKGROUND

The scientific literature contains millions of microbial gene identifiers within the full text and tables, but these annotations rarely get incorporated into public sequence databases. We propose to utilize the Open Access (OA) subset of PubMed Central (PMC) as a gene annotation database and have developed an R package called pmcXML to automatically mine and extract locus tags from full text, tables and supplements.

RESULTS

We mined locus tags from 1835 OA publications in ten microbial genomes and extracted tags mentioned in 30,891 sentences in main text and 20,489 rows in tables. We identified locus tag pairs marking the start and end of a region such as an operon or genomic island and expanded these ranges to add another 13,043 tags. We also searched for locus tags in supplementary tables and publications outside the OA subset in Burkholderia pseudomallei K96243 for comparison. There were 168 publications containing 48,470 locus tags and 83% of mentions were from supplementary materials and 9% from publications outside the OA subset.

CONCLUSIONS

B. pseudomallei locus tags within the full text and tables of OA publications represent only a small fraction of the total mentions in the literature. For microbial genomes with very few functionally characterized proteins, the locus tags mentioned in supplementary tables and within ranges like genomic islands contain the majority of locus tags. Significantly, the functions in the R package provide access to additional resources in the OA subset that are not currently indexed or returned by searching PMC.

Collective instance-level gene normalization on the IGN corpus

A high proportion of life science researches are gene-oriented, in which scientists aim to investigate the roles that genes play in biological processes, and their involvement in biological mechanisms. As a result, gene names and their related information turn out to be one of the main objects of interest in biomedical literatures. While the capability of recognizing gene mentions has made significant progress, the results of recognition are still insufficient for direct use due to the ambiguity of gene names. Gene normalization (GN) goes beyond the recognition task by linking a gene mention to a database ID. Unlike most previous works, we approach GN on the instance-level and evaluate its overall performance on the recognition and normalization steps in abstracts and full texts. We release the first instance-level gene normalization (IGN) corpus in the BioC format, which includes annotations for the boundaries of all gene mentions and the corresponding IDs for human gene mentions. Species information, along with existing co-reference chains and full name/abbreviation pairs are also provided for each gene mention. Using the released corpus, we have designed a collective instance-level GN approach using not only the contextual information of each individual instance, but also the relations among instances and the inherent characteristics of full-text sections. Our experimental results show that our collective approach can achieve an F-score of 0.743. The proposed approach that exploits section characteristics in full-text articles can improve the F-scores of information lacking sections by up to 1.8%. In addition, using the proposed refinement process improved the F-score of gene mention recognition by 0.125 and that of GN by 0.03. Whereas current experimental results are limited to the human species, we seek to continue updating the annotations of the IGN corpus and observe how the proposed approach can be extended to other species.

Wiki-pi: a web-server of annotated human protein-protein interactions to aid in discovery of protein function

Protein-protein interactions (PPIs) are the basis of biological functions. Knowledge of the interactions of a protein can help understand its molecular function and its association with different biological processes and pathways. Several publicly available databases provide comprehensive information about individual proteins, such as their sequence, structure, and function. There also exist databases that are built exclusively to provide PPIs by curating them from published literature. The information provided in these web resources is protein-centric, and not PPI-centric. The PPIs are typically provided as lists of interactions of a given gene with links to interacting partners; they do not present a comprehensive view of the nature of both the proteins involved in the interactions. A web database that allows search and retrieval based on biomedical characteristics of PPIs is lacking, and is needed. We present Wiki-Pi (read Wiki-π), a web-based interface to a database of human PPIs, which allows users to retrieve interactions by their biomedical attributes such as their association to diseases, pathways, drugs and biological functions. Each retrieved PPI is shown with annotations of both of the participant proteins side-by-side, creating a basis to hypothesize the biological function facilitated by the interaction. Conceptually, it is a search engine for PPIs analogous to PubMed for scientific literature. Its usefulness in generating novel scientific hypotheses is demonstrated through the study of IGSF21, a little-known gene that was recently identified to be associated with diabetic retinopathy. Using Wiki-Pi, we infer that its association to diabetic retinopathy may be mediated through its interactions with the genes HSPB1, KRAS, TMSB4X and DGKD, and that it may be involved in cellular response to external stimuli, cytoskeletal organization and regulation of molecular activity. The website also provides a wiki-like capability allowing users to describe or discuss an interaction. Wiki-Pi is available publicly and freely at http://severus.dbmi.pitt.edu/wiki-pi/.

BioContext: an integrated text mining system for large-scale extraction and contextualization of biomolecular events

Motivation: Although the amount of data in biology is rapidly increasing, critical information for understanding biological events like phosphorylation or gene expression remains locked in the biomedical literature. Most current text mining (TM) approaches to extract information about biological events are focused on either limited-scale studies and/or abstracts, with data extracted lacking context and rarely available to support further research.

Results: Here we present BioContext, an integrated TM system which extracts, extends and integrates results from a number of tools performing entity recognition, biomolecular event extraction and contextualization. Application of our system to 10.9 million MEDLINE abstracts and 234 000 open-access full-text articles from PubMed Central yielded over 36 million mentions representing 11.4 million distinct events. Event participants included over 290 000 distinct genes/proteins that are mentioned more than 80 million times and linked where possible to Entrez Gene identifiers. Over a third of events contain contextual information such as the anatomical location of the event occurrence or whether the event is reported as negated or speculative.

Availability: The BioContext pipeline is available for download (under the BSD license) at http://www.biocontext.org, along with the extracted data which is also available for online browsing.

Contact:martin.gerner@gmail.com

Supplementary information:Supplementary data are available at Bioinformatics online.