From molecular mechanisms of prostate cancer to translational applications: based on multi-omics fusion analysis and intelligent medicine

Prostate cancer is the most common cancer in men worldwide and has a high mortality rate. The complex and heterogeneous development of prostate cancer has become a core obstacle in the treatment of prostate cancer. Simultaneously, the issues of overtreatment in early-stage diagnosis, oligometastasis and dormant tumor recognition, as well as personalized drug utilization, are also specific concerns that require attention in the clinical management of prostate cancer. Some typical genetic mutations have been proved to be associated with prostate cancer's initiation and progression. However, single-omic studies usually are not able to explain the causal relationship between molecular alterations and clinical phenotypes. Exploration from a systems genetics perspective is also lacking in this field, that is, the impact of gene network, the environmental factors, and even lifestyle behaviors on disease progression. At the meantime, current trend emphasizes the utilization of artificial intelligence (AI) and machine learning techniques to process extensive multidimensional data, including multi-omics. These technologies unveil the potential patterns, correlations, and insights related to diseases, thereby aiding the interpretable clinical decision making and applications, namely intelligent medicine. Therefore, there is a pressing need to integrate multidimensional data for identification of molecular subtypes, prediction of cancer progression and aggressiveness, along with perosonalized treatment performing. In this review, we systematically elaborated the landscape from molecular mechanism discovery of prostate cancer to clinical translational applications. We discussed the molecular profiles and clinical manifestations of prostate cancer heterogeneity, the identification of different states of prostate cancer, as well as corresponding precision medicine practices. Taking multi-omics fusion, systems genetics, and intelligence medicine as the main perspectives, the current research results and knowledge-driven research path of prostate cancer were summarized.

CD34+ stromal cells/telocytes and their role in mouse lung development: Light microscopy, immunofluorescence, ultrastructural and scanning electron microscopy evidence

Telocytes (TCs), a novel type of mesenchymal or interstitial cell with specific, very long and thin cellular prolongations, have been found in various mammalian organs and have potential biological functions. However, their existence during lung development is poorly understood. This study aimed to investigate the existence, morphological features, and role of CD34+ SCs/TCs in mouse lungs from foetal to postnatal life using primary cell culture, double immunofluorescence, transmission electron microscopy (TEM) and scanning electron microscopy (SEM). The immunofluorescence double staining profiles revealed positive expression of CD34 and PDGFR-α, Sca-1 or VEGFR-3, and the expression of these markers differed among the age groups during lung development. Intriguingly, in the E18.5 stage of development, along with the CD34+ SCs/TCs, haematopoietic stem cells and angiogenic factors were also significantly increased in number compared with those in the E14.5, E16.5, P0 and P7. Subsequently, TEM confirmed that CD34+ SCs/TCs consisted of a small cell body with long telopodes (Tps) that projected from the cytoplasm. Tps consisted of alternating thin and thick segments known as podomers and podoms. TCs contain abundant endoplasmic reticulum, mitochondria and secretory vesicles and establish close connections with neighbouring cells. Furthermore, SEM revealed characteristic features, including triangular, oval, spherical, or fusiform cell bodies with extensive cellular prolongations, depending on the number of Tps. Our findings provide evidence for the existence of CD34+ SCs/TCs, which contribute to vasculogenesis, the formation of the air‒blood barrier, tissue organization during lung development and homoeostasis.

In search of the ratio of miRNA expression as robust biomarkers for constructing stable diagnostic models among multi-center data

MicroRNAs (miRNAs) are promising biomarkers for the early detection of disease, and many miRNA-based diagnostic models have been constructed to distinguish patients and healthy individuals. To thoroughly utilize the miRNA-profiling data across different sequencing platforms or multiple centers, the models accounting the batch effects were demanded for the generalization of medical application. We conducted transcription factor (TF)-mediated miRNA-miRNA interaction network analysis and adopted the within-sample expression ratios of miRNA pairs as predictive markers. The ratio of the expression values between each miRNA pair turned out to be stable across multiple data sources. A genetic algorithm-based classifier was constructed to quantify risk scores of the probability of disease and discriminate disease states from normal states in discovery, with a validation dataset for COVID-19, renal cell carcinoma, and lung adenocarcinoma. The predictive models based on the expression ratio of interacting miRNA pairs demonstrated good performances in the discovery and validation datasets, and the classifier may be used accurately for the early detection of disease.

Edge-based relative entropy as a sensitive indicator of critical transitions in biological systems

BACKGROUND

Disease progression in biosystems is not always a steady process but is occasionally abrupt. It is important but challenging to signal critical transitions in complex biosystems.

METHODS

In this study, based on the theoretical framework of dynamic network biomarkers (DNBs), we propose a model-free method, edge-based relative entropy (ERE), to identify temporal key biomolecular associations/networks that may serve as DNBs and detect early-warning signals of the drastic state transition during disease progression in complex biological systems. Specifically, by combining gene‒gene interaction (edge) information with the relative entropy, the ERE method converts gene expression values into network entropy values, quantifying the dynamic change in a biomolecular network and indicating the qualitative shift in the system state.

RESULTS

The proposed method was validated using simulated data and real biological datasets of complex diseases. The applications show that for certain diseases, the ERE method helps to reveal so-called "dark genes" that are non-differentially expressed but with high ERE values and of essential importance in both gene regulation and prognosis.

CONCLUSIONS

The proposed method effectively identified the critical transition states of complex diseases at the network level. Our study not only identified the critical transition states of various cancers but also provided two types of new prognostic biomarkers, positive and negative edge biomarkers, for further practical application. The method in this study therefore has great potential in personalized disease diagnosis.

Latent space search based multimodal optimization with personalized edge-network biomarker for multi-purpose early disease prediction

Considering that cancer is resulting from the comutation of several essential genes of individual patients, researchers have begun to focus on identifying personalized edge-network biomarkers (PEBs) using personalized edge-network analysis for clinical practice. However, most of existing methods ignored the optimization of PEBs when multimodal biomarkers exist in multi-purpose early disease prediction (MPEDP). To solve this problem, this study proposes a novel model (MMPDENB-RBM) that combines personalized dynamic edge-network biomarkers (PDENB) theory, multimodal optimization strategy and latent space search scheme to identify biomarkers with different configurations of PDENB modules (i.e. to effectively identify multimodal PDENBs). The application to the three largest cancer omics datasets from The Cancer Genome Atlas database (i.e. breast invasive carcinoma, lung squamous cell carcinoma and lung adenocarcinoma) showed that the MMPDENB-RBM model could more effectively predict critical cancer state compared with other advanced methods. And, our model had better convergence, diversity and multimodal property as well as effective optimization ability compared with the other state-of-art methods. Particularly, multimodal PDENBs identified were more enriched with different functional biomarkers simultaneously, such as tissue-specific synthetic lethality edge-biomarkers including cancer driver genes and disease marker genes. Importantly, as our aim, these multimodal biomarkers can perform diverse biological and biomedical significances for drug target screen, survival risk assessment and novel biomedical sight as the expected multi-purpose of personalized early disease prediction. In summary, the present study provides multimodal property of PDENBs, especially the therapeutic biomarkers with more biological significances, which can help with MPEDP of individual cancer patients.

Noninvasive detection and interpretation of gastrointestinal diseases by collaborative serum metabolite and magnetically controlled capsule endoscopy

Gastrointestinal diseases are complex diseases that occur in the gastrointestinal tract. Common gastrointestinal diseases include chronic gastritis, peptic ulcers, inflammatory bowel disease, and gastrointestinal tumors. These diseases may manifest a long course, difficult treatment, and repeated attacks.

Gastroscopy and mucosal biopsy are the gold standard methods for diagnosing gastric and duodenal diseases, but they are invasive procedures and carry risks due to the necessity of sedation and anesthesia. Recently, several new approaches have been developed, including serological examination and magnetically controlled capsule endoscopy (MGCE). However, serological markers lack lesion information, while MGCE images lack molecular information.

This study proposes combining these two technologies in a collaborative noninvasive diagnostic scheme as an alternative to the standard procedures. We introduce an interpretable framework for the clinical diagnosis of gastrointestinal diseases. Based on collected blood samples and MGCE records of patients with gastrointestinal diseases and comparisons with normal individuals, we selected serum metabolite signatures by bioinformatic analysis, captured image embedding signatures by convolutional neural networks, and inferred the location-specific associations between these signatures.

Our study successfully identified five key metabolite signatures with functional relevance to gastrointestinal disease. The combined signatures achieved discrimination AUC of 0.88. Meanwhile, the image embedding signatures showed different levels of validation and testing accuracy ranging from 0.7 to 0.9 according to different locations in the gastrointestinal tract as explained by their specific associations with metabolite signatures. Overall, our work provides a new collaborative noninvasive identification pipeline and candidate metabolite biomarkers for image auxiliary diagnosis. This method should be valuable for the noninvasive detection and interpretation of gastrointestinal and other complex diseases.

Network control principles for identifying personalized driver genes in cancer

To understand tumor heterogeneity in cancer, personalized driver genes (PDGs) need to be identified for unraveling the genotype-phenotype associations corresponding to particular patients. However, most of the existing driver-focus methods mainly pay attention on the cohort information rather than on individual information. Recent developing computational approaches based on network control principles are opening a new way to discover driver genes in cancer, particularly at an individual level. To provide comprehensive perspectives of network control methods on this timely topic, we first considered the cancer progression as a network control problem, in which the expected PDGs are altered genes by oncogene activation signals that can change the individual molecular network from one health state to the other disease state. Then, we reviewed the network reconstruction methods on single samples and introduced novel network control methods on single-sample networks to identify PDGs in cancer. Particularly, we gave a performance assessment of the network structure control-based PDGs identification methods on multiple cancer datasets from TCGA, for which the data and evaluation package also are publicly available. Finally, we discussed future directions for the application of network control methods to identify PDGs in cancer and diverse biological processes.

Identification of Autistic Risk Candidate Genes and Toxic Chemicals via Multilabel Learning

As a group of complex neurodevelopmental disorders, autism spectrum disorder (ASD) has been reported to have a high overall prevalence, showing an unprecedented spurt since 2000. Due to the unclear pathomechanism of ASD, it is challenging to diagnose individuals with ASD merely based on clinical observations. Without additional support of biochemical markers, the difficulty of diagnosis could impact therapeutic decisions and, therefore, lead to delayed treatments. Recently, accumulating evidence have shown that both genetic abnormalities and chemical toxicants play important roles in the onset of ASD. In this work, a new multilabel classification (MLC) model is proposed to identify the autistic risk genes and toxic chemicals on a large-scale data set. We first construct the feature matrices and partially labeled networks for autistic risk genes and toxic chemicals from multiple heterogeneous biological databases. Based on both global and local measure metrics, the simulation experiments demonstrate that the proposed model achieves superior classification performance in comparison with the other state-of-the-art MLC methods. Through manual validation with existing studies, 60% and 50% out of the top-20 predicted risk genes are confirmed to have associations with ASD and autistic disorder, respectively. To the best of our knowledge, this is the first computational tool to identify ASD-related risk genes and toxic chemicals, which could lead to better therapeutic decisions of ASD.

Temporal discounting as a candidate behavioral marker of obesity

Although obesity is a result of processes operating at multiple levels, most forms result from decision-making behavior. The aim of this review was to examine the candidacy of temporal discounting (TD) (i.e. the reduction in the value of a reinforcer as a function of the delay to its receipt) as a behavioral marker of obesity. For this purpose, we assessed whether TD has the ability to: identify risk for obesity development, diagnose obesity, track obesity progression, predict treatment prognosis/outcomes, and measure treatment effectiveness. Three databases (Pubmed, PsycINFO, and Web of Science) were searched using a combination of terms related to TD and obesity. A total of 153 papers were reviewed. Several areas show strong evidence of TD's predictive utility as a behavioral marker of obesity (e.g., distinguishing obese from non obese). However, other areas have limited and/or mixed evidence (e.g., predicting weight change). Given the positive relationship for TD in the majority of domains examined, further consideration for TD as a behavioral marker of obesity is warranted.

Network controllability-based algorithm to target personalized driver genes for discovering combinatorial drugs of individual patients

Multiple driver genes in individual patient samples may cause resistance to individual drugs in precision medicine. However, current computational methods have not studied how to fill the gap between personalized driver gene identification and combinatorial drug discovery for individual patients. Here, we developed a novel structural network controllability-based personalized driver genes and combinatorial drug identification algorithm (CPGD), aiming to identify combinatorial drugs for an individual patient by targeting personalized driver genes from network controllability perspective. On two benchmark disease datasets (i.e. breast cancer and lung cancer datasets), performance of CPGD is superior to that of other state-of-the-art driver gene-focus methods in terms of discovery rate among prior-known clinical efficacious combinatorial drugs. Especially on breast cancer dataset, CPGD evaluated synergistic effect of pairwise drug combinations by measuring synergistic effect of their corresponding personalized driver gene modules, which are affected by a given targeting personalized driver gene set of drugs. The results showed that CPGD performs better than existing synergistic combinatorial strategies in identifying clinical efficacious paired combinatorial drugs. Furthermore, CPGD enhanced cancer subtyping by computationally providing personalized side effect signatures for individual patients. In addition, CPGD identified 90 drug combinations candidates from SARS-COV2 dataset as potential drug repurposing candidates for recently spreading COVID-19.

Applying artificial intelligence in the microbiome for gastrointestinal diseases: A review

For a long time, gut bacteria have been recognized for their important roles in the occurrence and progression of gastrointestinal diseases like colorectal cancer, and the ever-increasing amounts of microbiome data combined with other high-quality clinical and imaging datasets are leading the study of gastrointestinal diseases into an era of biomedical big data. The "omics" technologies used for microbiome analysis continuously evolve, and the machine learning or artificial intelligence technologies are key to extract the relevant information from microbiome data. This review intends to provide a focused summary of recent research and applications of microbiome big data and to discuss the use of artificial intelligence to combat gastrointestinal diseases.

Dynamic edge-based biomarker non-invasively predicts hepatocellular carcinoma with hepatitis B virus infection for individual patients based on blood testing

Hepatitis B virus (HBV)-induced hepatocellular carcinoma (HCC) is a major cause of cancer-related deaths in Asia and Africa. Developing effective and non-invasive biomarkers of HCC for individual patients remains an urgent task for early diagnosis and convenient monitoring. Analyzing the transcriptomic profiles of peripheral blood mononuclear cells from both healthy donors and patients with chronic HBV infection in different states (i.e. HBV carrier, chronic hepatitis B, cirrhosis, and HCC), we identified a set of 19 candidate genes according to our algorithm of dynamic network biomarkers. These genes can both characterize different stages during HCC progression and identify cirrhosis as the critical transition stage before carcinogenesis. The interaction effects (i.e. co-expressions) of candidate genes were used to build an accurate prediction model: the so-called edge-based biomarker. Considering the convenience and robustness of biomarkers in clinical applications, we performed functional analysis, validated candidate genes in other independent samples of our collected cohort, and finally selected COL5A1, HLA-DQB1, MMP2, and CDK4 to build edge panel as prediction models. We demonstrated that the edge panel had great performance in both diagnosis and prognosis in terms of precision and specificity for HCC, especially for patients with alpha-fetoprotein-negative HCC. Our study not only provides a novel edge-based biomarker for non-invasive and effective diagnosis of HBV-associated HCC to each individual patient but also introduces a new way to integrate the interaction terms of individual molecules for clinical diagnosis and prognosis from the network and dynamics perspectives.

Rewiring of Microbiota Networks in Erosive Inflammation of the Stomach and Small Bowel

The development of non-invasive, inexpensive, and effective early diagnosis tests for gastric and small-bowel lesions is an urgent requirement. The introduction of magnetically guided capsule endoscopy (MGCE) has aided examination of the small bowel for diagnoses. However, the distribution of the fecal microbiome in abnormal erosions of the stomach and small bowel remains unclear. Herein, alternations in the fecal microbiome in three groups [normal, small-bowel inflammation, and chronic gastritis (CG)] were analyzed by metagenomics and our well-developed method [individual-specific edge-network analysis (iENA)]. In addition to the dominant microbiota identified by the conventional differential analysis, iENA could recognize novel network biomarkers of microbiome communities, such as the genus Bacteroide in CG and small-bowel inflammation. Combined with differential network analysis, the network-hub microbiota within rewired microbiota networks revealed high-ranked iENA microbiota markers, which were disease specific and had particular pathogenic functions. Our findings illuminate the components of the fecal microbiome and the importance of specific bacteria in CG and small-bowel erosions, and could be employed to develop preventive and non-invasive therapeutic strategies.

Computer-aided biomarker discovery for precision medicine: data resources, models and applications

Biomarkers are a class of measurable and evaluable indicators with the potential to predict disease initiation and progression. In contrast to disease-associated factors, biomarkers hold the promise to capture the changeable signatures of biological states. With methodological advances, computer-aided biomarker discovery has now become a burgeoning paradigm in the field of biomedical science. In recent years, the 'big data' term has accumulated for the systematical investigation of complex biological phenomena and promoted the flourishing of computational methods for systems-level biomarker screening. Compared with routine wet-lab experiments, bioinformatics approaches are more efficient to decode disease pathogenesis under a holistic framework, which is propitious to identify biomarkers ranging from single molecules to molecular networks for disease diagnosis, prognosis and therapy. In this review, the concept and characteristics of typical biomarker types, e.g. single molecular biomarkers, module/network biomarkers, cross-level biomarkers, etc., are explicated on the guidance of systems biology. Then, publicly available data resources together with some well-constructed biomarker databases and knowledge bases are introduced. Biomarker identification models using mathematical, network and machine learning theories are sequentially discussed. Based on network substructural and functional evidences, a novel bioinformatics model is particularly highlighted for microRNA biomarker discovery. This article aims to give deep insights into the advantages and challenges of current computational approaches for biomarker detection, and to light up the future wisdom toward precision medicine and nation-wide healthcare.

The detection and analysis of differential regulatory communities in lung cancer

The tumorgenesis process of lung cancer involves the regulatory dysfunctions of multiple pathways. Although many signaling pathways have been identified to be associated with lung cancer, there are little quantitative models of how inactions between genes change during the process from normal to cancer. These changes belong to different dynamic co-expressions patterns. We quantitatively analyzed differential co-expression of gene pairs in four datasets. Each dataset included a large number of lung cancer and normal samples. By overlapping their results, we got 14 highly confident gene pairs with consistent co-expression change patterns. Some of they, such as ARHGAP30 and GIMAP4, had been recorded in STRING network database while some of them were novel discoveries, such as C9orf135 and MORN5, TEKT1 and TSPAN1 were positively correlated in both normal and cancer but more correlated in normal than cancer. These gene pairs revealed the underlying mechanisms of lung cancer occurrence.

Roles of transforming growth factor-β and phosphatidylinositol 3-kinase isoforms in integrin β1-mediated bio-behaviors of mouse lung telocytes

Background

Telocytes (TCs) have the capacity of cell–cell communication with adjacent cells within the tissue, contributing to tissue repair and recovery from injury. The present study aims at investigating the molecular mechanisms by which the TGFβ1-ITGB1-PI3K signal pathways regulate TC cycle and proliferation.

Methods

Gene expression of integrin (ITG) family were measured in mouse primary TCs to compare with other cells. TC proliferation, movement, cell cycle, and PI3K isoform protein genes were assayed in ITGB1-negative or positive mouse lung TCs treated with the inhibition of PI3Kp110α, PI3Kα/δ, PKCβ, or GSK3, followed by TGFβ1 treatment.

Results

We found the characters and interactions of ITG or PKC family member networks in primary mouse lung TCs, different from other cells in the lung tissue. The deletion of ITGB1 changed TCs sensitivity to treatment with multifunctional cytokines or signal pathway inhibitors. The compensatory mechanisms occur among TGFβ1-induced PI3Kp110α, PI3Kα/δ, PKCβ, or GSK3 when ITGB1 gene was deleted, leading to alterations of TC cell cycle and proliferation. Of those PI3K isoform protein genes, mRNA expression of PIK3CG altered with ITGB1-negative TC cycle and proliferation.

Conclusion

TCs have strong capacity of proliferation through the compensatory signaling mechanisms and contribute to the development of drug resistance due to alterations of TC sensitivity.

Dynamically characterizing individual clinical change by the steady state of disease-associated pathway

Background

Along with the development of precision medicine, individual heterogeneity is attracting more and more attentions in clinical research and application. Although the biomolecular reaction seems to be some various when different individuals suffer a same disease (e.g. virus infection), the final pathogen outcomes of individuals always can be mainly described by two categories in clinics, i.e. symptomatic and asymptomatic. Thus, it is still a great challenge to characterize the individual specific intrinsic regulatory convergence during dynamic gene regulation and expression. Except for individual heterogeneity, the sampling time also increase the expression diversity, so that, the capture of similar steady biological state is a key to characterize individual dynamic biological processes.

Results

Assuming the similar biological functions (e.g. pathways) should be suitable to detect consistent functions rather than chaotic genes, we design and implement a new computational framework (ABP: Attractor analysis of Boolean network of Pathway). ABP aims to identify the dynamic phenotype associated pathways in a state-transition manner, using the network attractor to model and quantify the steady pathway states characterizing the final steady biological sate of individuals (e.g. normal or disease). By analyzing multiple temporal gene expression datasets of virus infections, ABP has shown its effectiveness on identifying key pathways associated with phenotype change; inferring the consensus functional cascade among key pathways; and grouping pathway activity states corresponding to disease states.

Conclusions

Collectively, ABP can detect key pathways and infer their consensus functional cascade during dynamical process (e.g. virus infection), and can also categorize individuals with disease state well, which is helpful for disease classification and prediction.

Decoding the similarities and specific differences between latent and active tuberculosis infections based on consistently differential expression networks

Although intensive efforts have been devoted to investigating latent tuberculosis (LTB) and active tuberculosis (PTB) infections, the similarities and differences in the host responses to these two closely associated stages remain elusive, probably due to the difficulty in identifying informative genes related to LTB using traditional methods. Herein, we developed a framework known as the consistently differential expression network to identify tuberculosis (TB)-related gene pairs by combining microarray profiles and protein-protein interactions. We thus obtained 774 and 693 pairs corresponding to the PTB and LTB stages, respectively. The PTB-specific genes showed higher expression values and fold-changes than the LTB-specific genes. Furthermore, the PTB-related pairs generally had higher expression correlations and would be more activated compared to their LTB-related counterparts. The module analysis implied that the detected gene pairs tended to cluster in the topological and functional modules. Functional analysis indicated that the LTB- and PTB-specific genes were enriched in different pathways and had remarkably different locations in the NF-κB signaling pathway. Finally, we showed that the identified genes and gene pairs had the potential to distinguish TB patients in different disease stages and could be considered as drug targets for the specific treatment of patients with LTB or PTB.

Feature selection may improve deep neural networks for the bioinformatics problems

Motivation

Deep neural network (DNN) algorithms were utilized in predicting various biomedical phenotypes recently, and demonstrated very good prediction performances without selecting features. This study proposed a hypothesis that the DNN models may be further improved by feature selection algorithms.

Results

A comprehensive comparative study was carried out by evaluating 11 feature selection algorithms on three conventional DNN algorithms, i.e. convolution neural network (CNN), deep belief network (DBN) and recurrent neural network (RNN), and three recent DNNs, i.e. MobilenetV2, ShufflenetV2 and Squeezenet. Five binary classification methylomic datasets were chosen to calculate the prediction performances of CNN/DBN/RNN models using feature selected by the 11 feature selection algorithms. Seventeen binary classification transcriptome and two multi-class transcriptome datasets were also utilized to evaluate how the hypothesis may generalize to different data types. The experimental data supported our hypothesis that feature selection algorithms may improve DNN models, and the DBN models using features selected by SVM-RFE usually achieved the best prediction accuracies on the five methylomic datasets.

Availability and implementation

All the algorithms were implemented and tested under the programming environment Python version 3.6.6.

Supplementary information

Supplementary data are available at Bioinformatics online.

Roles of TGFβ1 in the expression of phosphoinositide 3-kinase isoform genes and sensitivity and response of lung telocytes to PI3K inhibitors

BACKGROUND

The mouse lung telocyte cell line (TC^SV40) recently established provides further opportunities to learn TC biology and functions. The present study aims at investigating regulatory roles of phosphoinositide 3-kinase (PI3K) isoforms in TC proliferation and movement and in TGFβ1-induced sensitivity and response of lung TCs to PI3K inhibitors.

MATERIALS AND METHODS

Network and molecular interactions of genes coding PI3K family or TGFβ family proteins in mouse primary TCs were defined. Mouse lung TC^SV40 proliferation, apoptosis, cell cycle, and dynamical bio-behaviors were measured with or without TGFβ1 stimulation or PI3K catalytic isoform protein (PI3K/mTOR, PI3Kα/δ/β, PI3K p110δ, or pan-PI3K) inhibitions.

RESULTS

The present study showed the difference of network characteristics and interactions of genes coding PI3K isoform proteins or TGFβ family proteins in primary lung telocytes from mouse lungs compared to those of other cells residing in the lung. TGFβ1 had diverse effects on TC proliferation with altered TC number in G2 or S phase, independent upon the administered dose of TGFβ1. PI3Kα/δ/β, PI3K/mTOR, and PI3K p110δ were involved in TC proliferation, of which PI3Kα/δ/β was more sensitive. The effects of pan-PI3K inhibitor indicate that more PI3K isoforms were stimulated by the administering of external TGFβ1 and contributed to TGFβ1-induced TC proliferation. PI3K p110δ upregulated TC proliferation and movement dynamically without TGFβ1, and downregulated TC proliferation with TGFβ1 stimulation, but not TC movement. PI3Kα/δ/β and PI3K/mTOR were more active in TGFβ1-induced S phase accumulation and had similar dynamic effects to PI3K p110δ. Gene expression of PI3K isoforms in TCs was upregulated after TGFβ1 stimulation. The expression of PIK3CA coding p110-α or PIK3CG coding p110-γ were up- or downregulated in TCs without TGFβ1, respectively, when PI3K/mTOR, PI3Kα/δ/β, PI3K p110δ, or pan-PI3K were inhibited. TGFβ1 upregulated the expression of PIK3CA and PIK3CB, while downregulated the expression of PIK3CD and PIK3CG.

CONCLUSION

Our data imply that TGFβ1 plays divergent roles in the expression of PI3K isoform genes in lung TCs and can alter the sensitivity and response of lung TCs to PI3K inhibitors.

Single-Cell RNA Sequencing-Based Computational Analysis to Describe Disease Heterogeneity

The trillions of cells in the human body can be viewed as elementary but essential biological units that achieve different body states, but the low resolution of previous cell isolation and measurement approaches limits our understanding of the cell-specific molecular profiles. The recent establishment and rapid growth of single-cell sequencing technology has facilitated the identification of molecular profiles of heterogeneous cells, especially on the transcription level of single cells [single-cell RNA sequencing (scRNA-seq)]. As a novel method, the robustness of scRNA-seq under changing conditions will determine its practical potential in major research programs and clinical applications. In this review, we first briefly presented the scRNA-seq-related methods from the point of view of experiments and computation. Then, we compared several state-of-the-art scRNA-seq analysis frameworks mainly by analyzing their performance robustness on independent scRNA-seq datasets for the same complex disease. Finally, we elaborated on our hypothesis on consensus scRNA-seq analysis and summarized the potential indicative and predictive roles of individual cells in understanding disease heterogeneity by single-cell technologies.

Concepts of Artificial Intelligence for Computer-Assisted Drug Discovery

Artificial intelligence (AI), and, in particular, deep learning as a subcategory of AI, provides opportunities for the discovery and development of innovative drugs. Various machine learning approaches have recently (re)emerged, some of which may be considered instances of domain-specific AI which have been successfully employed for drug discovery and design. This review provides a comprehensive portrayal of these machine learning techniques and of their applications in medicinal chemistry. After introducing the basic principles, alongside some application notes, of the various machine learning algorithms, the current state-of-the art of AI-assisted pharmaceutical discovery is discussed, including applications in structure- and ligand-based virtual screening, de novo drug design, physicochemical and pharmacokinetic property prediction, drug repurposing, and related aspects. Finally, several challenges and limitations of the current methods are summarized, with a view to potential future directions for AI-assisted drug discovery and design.

Mathematical modeling and computational prediction of cancer drug resistance

Diverse forms of resistance to anticancer drugs can lead to the failure of chemotherapy. Drug resistance is one of the most intractable issues for successfully treating cancer in current clinical practice. Effective clinical approaches that could counter drug resistance by restoring the sensitivity of tumors to the targeted agents are urgently needed. As numerous experimental results on resistance mechanisms have been obtained and a mass of high-throughput data has been accumulated, mathematical modeling and computational predictions using systematic and quantitative approaches have become increasingly important, as they can potentially provide deeper insights into resistance mechanisms, generate novel hypotheses or suggest promising treatment strategies for future testing. In this review, we first briefly summarize the current progress of experimentally revealed resistance mechanisms of targeted therapy, including genetic mechanisms, epigenetic mechanisms, posttranslational mechanisms, cellular mechanisms, microenvironmental mechanisms and pharmacokinetic mechanisms. Subsequently, we list several currently available databases and Web-based tools related to drug sensitivity and resistance. Then, we focus primarily on introducing some state-of-the-art computational methods used in drug resistance studies, including mechanism-based mathematical modeling approaches (e.g. molecular dynamics simulation, kinetic model of molecular networks, ordinary differential equation model of cellular dynamics, stochastic model, partial differential equation model, agent-based model, pharmacokinetic–pharmacodynamic model, etc.) and data-driven prediction methods (e.g. omics data-based conventional screening approach for node biomarkers, static network approach for edge biomarkers and module biomarkers, dynamic network approach for dynamic network biomarkers and dynamic module network biomarkers, etc.). Finally, we discuss several further questions and future directions for the use of computational methods for studying drug resistance, including inferring drug-induced signaling networks, multiscale modeling, drug combinations and precision medicine.

Dysfunction of PLA2G6 and CYP2C44-associated network signals imminent carcinogenesis from chronic inflammation to hepatocellular carcinoma

Little is known about how chronic inflammation contributes to the progression of hepatocellular carcinoma (HCC), especially the initiation of cancer. To uncover the critical transition from chronic inflammation to HCC and the molecular mechanisms at a network level, we analyzed the time-series proteomic data of woodchuck hepatitis virus/c-myc mice and age-matched wt-C57BL/6 mice using our dynamical network biomarker (DNB) model. DNB analysis indicated that the 5th month after birth of transgenic mice was the critical period of cancer initiation, just before the critical transition, which is consistent with clinical symptoms. Meanwhile, the DNB-associated network showed a drastic inversion of protein expression and coexpression levels before and after the critical transition. Two members of DNB, PLA2G6 and CYP2C44, along with their associated differentially expressed proteins, were found to induce dysfunction of arachidonic acid metabolism, further activate inflammatory responses through inflammatory mediator regulation of transient receptor potential channels, and finally lead to impairments of liver detoxification and malignant transition to cancer. As a c-Myc target, PLA2G6 positively correlated with c-Myc in expression, showing a trend from decreasing to increasing during carcinogenesis, with the minimal point at the critical transition or tipping point. Such trend of homologous PLA2G6 and c-Myc was also observed during human hepatocarcinogenesis, with the minimal point at high-grade dysplastic nodules (a stage just before the carcinogenesis). Our study implies that PLA2G6 might function as an oncogene like famous c-Myc during hepatocarcinogenesis, while downregulation of PLA2G6 and c-Myc could be a warning signal indicating imminent carcinogenesis.

Characterizing the Personalized Microbiota Dynamics for Disease Classification by Individual-Specific Edge-Network Analysis

Environmental factors such as the gut microbiome are thought to play an important role in the development and treatment of many diseases. But our understanding of microbiota compositional dynamics is still unclear and incomplete because the intestinal microbial community is an easily-changed ecosystem. It is urgently required to understand the large variations among individuals. These variations, however, will be an asset rather than a limitation to personalized medicine. For a proof-of-concept study on individual-specific disease classification based on microbiota compositional dynamics, we implemented an adjusted individual-specific edge-network analysis (iENA) method to address a limited number of samples from one individual, and compared it to the temporal 16S rRNA (ribosomal RNA) gene sequencing data from individuals in a challenge study. Our identified individual-specific OTU markers or their combined markers are consistent with previously reported markers, and the predictive score based on them can perform a better AUROC than the previous 0.83 and achieve about 90% accuracy on predicting whether an individual developed diarrhea [i.e., were symptomatic (Sx)] or not. In addition, iENA also showed satisfactory efficiency on another dataset about bacterial vaginosis (BV). All these results suggest that the combination of high-throughput microbiome experiments and computational systems biology approaches can efficiently recommend potential candidate species in the defense against various pathogens for precision medicine.

High-Order Correlation Integration for Single-Cell or Bulk RNA-seq Data Analysis

Quantifying or labeling the sample type with high quality is a challenging task, which is a key step for understanding complex diseases. Reducing noise pollution to data and ensuring the extracted intrinsic patterns in concordance with the primary data structure are important in sample clustering and classification. Here we propose an effective data integration framework named as HCI (High-order Correlation Integration), which takes an advantage of high-order correlation matrix incorporated with pattern fusion analysis (PFA), to realize high-dimensional data feature extraction. On the one hand, the high-order Pearson's correlation coefficient can highlight the latent patterns underlying noisy input datasets and thus improve the accuracy and robustness of the algorithms currently available for sample clustering. On the other hand, the PFA can identify intrinsic sample patterns efficiently from different input matrices by optimally adjusting the signal effects. To validate the effectiveness of our new method, we firstly applied HCI on four single-cell RNA-seq datasets to distinguish the cell types, and we found that HCI is capable of identifying the prior-known cell types of single-cell samples from scRNA-seq data with higher accuracy and robustness than other methods under different conditions. Secondly, we also integrated heterogonous omics data from TCGA datasets and GEO datasets including bulk RNA-seq data, which outperformed the other methods at identifying distinct cancer subtypes. Within an additional case study, we also constructed the mRNA-miRNA regulatory network of colorectal cancer based on the feature weight estimated from HCI, where the differentially expressed mRNAs and miRNAs were significantly enriched in well-known functional sets of colorectal cancer, such as KEGG pathways and IPA disease annotations. All these results supported that HCI has extensive flexibility and applicability on sample clustering with different types and organizations of RNA-seq data.

Systems Signatures Reveal Unique Remission-path of Type 2 Diabetes Following Roux-en-Y Gastric Bypass Surgery

Roux-en-Y Gastric bypass surgery (RYGB) is emerging as a powerful tool for treatment of obesity and may also cause remission of type 2 diabetes. However, the molecular mechanism of RYGB leading to diabetes remission independent of weight loss remains elusive. In this study, we profiled plasma metabolites and proteins of 10 normal glucose-tolerant obese (NO) and 9 diabetic obese (DO) patients before and 1-week, 3-months, 1-year after RYGB. 146 proteins and 128 metabolites from both NO and DO groups at all four stages were selected for further analysis. By analyzing a set of bi-molecular associations among the corresponding network of the subjects with our newly developed computational method, we defined the represented physiological states (called the edge-states that reflect the interactions among the bio-molecules), and the related molecular networks of NO and DO patients, respectively. The principal component analyses (PCA) revealed that the edge states of the post-RYGB NO subjects were significantly different from those of the post-RYGB DO patients. Particularly, the time-dependent changes of the molecular hub-networks differed between DO and NO groups after RYGB. In conclusion, by developing molecular network-based systems signatures, we for the first time reveal that RYGB generates a unique path for diabetes remission independent of weight loss.

A novel algorithm for finding optimal driver nodes to target control complex networks and its applications for drug targets identification

Background

The advances in target control of complex networks not only can offer new insights into the general control dynamics of complex systems, but also be useful for the practical application in systems biology, such as discovering new therapeutic targets for disease intervention. In many cases, e.g. drug target identification in biological networks, we usually require a target control on a subset of nodes (i.e., disease-associated genes) with minimum cost, and we further expect that more driver nodes consistent with a certain well-selected network nodes (i.e., prior-known drug-target genes).

Results

Therefore, motivated by this fact, we pose and address a new and practical problem called as target control problem with objectives-guided optimization (TCO): how could we control the interested variables (or targets) of a system with the optional driver nodes by minimizing the total quantity of drivers and meantime maximizing the quantity of constrained nodes among those drivers. Here, we design an efficient algorithm (TCOA) to find the optional driver nodes for controlling targets in complex networks. We apply our TCOA to several real-world networks, and the results support that our TCOA can identify more precise driver nodes than the existing control-fucus approaches. Furthermore, we have applied TCOA to two bimolecular expert-curate networks. Source code for our TCOA is freely available from http://sysbio.sibcb.ac.cn/cb/chenlab/software.htm or https://github.com/WilfongGuo/guoweifeng.

Conclusions

In the previous theoretical research for the full control, there exists an observation and conclusion that the driver nodes tend to be low-degree nodes. However, for target control the biological networks, we find interestingly that the driver nodes tend to be high-degree nodes, which is more consistent with the biological experimental observations. Furthermore, our results supply the novel insights into how we can efficiently target control a complex system, and especially many evidences on the practical strategic utility of TCOA to incorporate prior drug information into potential drug-target forecasts. Thus applicably, our method paves a novel and efficient way to identify the drug targets for leading the phenotype transitions of underlying biological networks.

Electronic supplementary material

The online version of this article (doi: 10.1186/s12864-017-4332-z) contains supplementary material, which is available to authorized users.

Individualized network-based drug repositioning infrastructure for precision oncology in the panomics era

Advances in next-generation sequencing technologies have generated the data supporting a large volume of somatic alterations in several national and international cancer genome projects, such as The Cancer Genome Atlas and the International Cancer Genome Consortium. These cancer genomics data have facilitated the revolution of a novel oncology drug discovery paradigm from candidate target or gene studies toward targeting clinically relevant driver mutations or molecular features for precision cancer therapy. This focuses on identifying the most appropriately targeted therapy to an individual patient harboring a particularly genetic profile or molecular feature. However, traditional experimental approaches that are used to develop new chemical entities for targeting the clinically relevant driver mutations are costly and high-risk. Drug repositioning, also known as drug repurposing, re-tasking or re-profiling, has been demonstrated as a promising strategy for drug discovery and development. Recently, computational techniques and methods have been proposed for oncology drug repositioning and identifying pharmacogenomics biomarkers, but overall progress remains to be seen. In this review, we focus on introducing new developments and advances of the individualized network-based drug repositioning approaches by targeting the clinically relevant driver events or molecular features derived from cancer panomics data for the development of precision oncology drug therapies (e.g. one-person trials) to fully realize the promise of precision medicine. We discuss several potential challenges (e.g. tumor heterogeneity and cancer subclones) for precision oncology. Finally, we highlight several new directions for the precision oncology drug discovery via biotherapies (e.g. gene therapy and immunotherapy) that target the 'undruggable' cancer genome in the functional genomics era.

Individual-specific edge-network analysis for disease prediction

Predicting pre-disease state or tipping point just before irreversible deterioration of health is a difficult task. Edge-network analysis (ENA) with dynamic network biomarker (DNB) theory opens a new way to study this problem by exploring rich dynamical and high-dimensional information of omics data. Although theoretically ENA has the ability to identify the pre-disease state during the disease progression, it requires multiple samples for such prediction on each individual, which are generally not available in clinical practice, thus limiting its applications in personalized medicine. In this work to overcome this problem, we propose the individual-specific ENA (iENA) with DNB to identify the pre-disease state of each individual in a single-sample manner. In particular, iENA can identify individual-specific biomarkers for the disease prediction, in addition to the traditional disease diagnosis. To demonstrate the effectiveness, iENA was applied to the analysis on omics data of H3N2 cohorts and successfully detected early-warning signals of the influenza infection for each individual both on the occurred time and event in an accurate manner, which actually achieves the AUC larger than 0.9. iENA not only found the new individual-specific biomarkers but also recovered the common biomarkers of influenza infection reported from previous works. In addition, iENA also detected the critical stages of multiple cancers with significant edge-biomarkers, which were further validated by survival analysis on both TCGA data and other independent data.

Chromosomal and Genomic Variations in Esophageal Squamous Cell Carcinoma: A Review of Technologies, Applications, and Prospections

Esophageal squamous cell carcinoma (ESCC) is one of the most common malignant tumors with poor prognosis worldwide. The poor prognosis is due to the advanced stage at the time of diagnosis and the limited clinical staging lacking significant molecular biomarkers to effectively stratify patients for treatment options. As cancer is a disease of genome instability and a resulting of accumulation of genetic alteration, mounting chromosomal and genomic technologies were developed and progressed rapidly which could be used for characterizing patients in genomics level. In this review, we summarized applications of multiple technologies and research progress at chromosomal and genomic level in ESCC.

Integrative enrichment analysis: a new computational method to detect dysregulated pathways in heterogeneous samples

BACKGROUND

Pathway enrichment analysis is a useful tool to study biology and biomedicine, due to its functional screening on well-defined biological procedures rather than separate molecules. The measurement of malfunctions of pathways with a phenotype change, e.g., from normal to diseased, is the key issue when applying enrichment analysis on a pathway. The differentially expressed genes (DEGs) are widely focused in conventional analysis, which is based on the great purity of samples. However, the disease samples are usually heterogeneous, so that, the genes with great differential expression variance (DEVGs) are becoming attractive and important to indicate the specific state of a biological system. In the context of differential expression variance, it is still a challenge to measure the enrichment or status of a pathway. To address this issue, we proposed Integrative Enrichment Analysis (IEA) based on a novel enrichment measurement.

RESULTS

The main competitive ability of IEA is to identify dysregulated pathways containing DEGs and DEVGs simultaneously, which are usually under-scored by other methods. Next, IEA provides two additional assistant approaches to investigate such dysregulated pathways. One is to infer the association among identified dysregulated pathways and expected target pathways by estimating pathway crosstalks. The other one is to recognize subtype-factors as dysregulated pathways associated to particular clinical indices according to the DEVGs' relative expressions rather than conventional raw expressions. Based on a previously established evaluation scheme, we found that, in particular cohorts (i.e., a group of real gene expression datasets from human patients), a few target disease pathways can be significantly high-ranked by IEA, which is more effective than other state-of-the-art methods. Furthermore, we present a proof-of-concept study on Diabetes to indicate: IEA rather than conventional ORA or GSEA can capture the under-estimated dysregulated pathways full of DEVGs and DEGs; these newly identified pathways could be significantly linked to prior-known disease pathways by estimated crosstalks; and many candidate subtype-factors recognized by IEA also have significant relation with the risk of subtypes of genotype-phenotype associations.

CONCLUSIONS

Totally, IEA supplies a new tool to carry on enrichment analysis in the complicate context of clinical application (i.e., heterogeneity of disease), as a necessary complementary and cooperative approach to conventional ones.