Removing batch effects from purified plasma cell gene expression microarrays with modified ComBat

Review on radiomic analysis in 18F-fluorodeoxyglucose positron emission tomography for prediction of melanoma outcomes

Over the past decade, several strategies have revolutionized the clinical management of patients with cutaneous melanoma (CM), including immunotherapy and targeted tyrosine kinase inhibitor (TKI)-based therapies. Indeed, immune checkpoint inhibitors (ICIs), alone or in combination, represent the standard of care for patients with advanced disease without an actionable mutation. Notably BRAF combined with MEK inhibitors represent the therapeutic standard for disease disclosing BRAF mutation. At the same time, FDG PET/CT has become part of the routine staging and evaluation of patients with cutaneous melanoma. There is growing interest in using FDG PET/CT measurements to predict response to ICI therapy and/or target therapy. While semiquantitative values such as standardized uptake value (SUV) are limited for predicting outcome, new measures including tumor metabolic volume, total lesion glycolysis and radiomics seem promising as potential imaging biomarkers for nuclear medicine. The aim of this review, prepared by an interdisciplinary group of experts, is to take stock of the current literature on radiomics approaches that could improve outcomes in CM.

Identifying radiogenomic associations of breast cancer based on DCE-MRI by using Siamese Neural Network with manufacturer bias normalization

BACKGROUND AND PURPOSE

The immunohistochemical test (IHC) for Human Epidermal Growth Factor Receptor 2 (HER2) and hormone receptors (HR) provides prognostic information and guides treatment for patients with invasive breast cancer. The objective of this paper is to establish a non-invasive system for identifying HER2 and HR in breast cancer using dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI).

METHODS

In light of the absence of high-performance algorithms and external validation in previously published methods, this study utilizes 3D deep features and radiomics features to represent the information of the Region of Interest (ROI). A Siamese Neural Network was employed as the classifier, with 3D deep features and radiomics features serving as the network input. To neutralize manufacturer bias, a batch effect normalization method, ComBat, was introduced. To enhance the reliability of the study, two datasets, Predict Your Therapeutic Response with Imaging and moLecular Analysis (I-SPY 1) and I-SPY 2, were incorporated. I-SPY 2 was utilized for model training and validation, while I-SPY 1 was exclusively employed for external validation. Additionally, a breast tumor segmentation network was trained to improve radiomic feature extraction.

RESULTS

The results indicate that our approach achieved an average Area Under the Curve (AUC) of 0.632, with a Standard Error of the Mean (SEM) of 0.042 for HER2 prediction in the I-SPY 2 dataset. For HR prediction, our method attained an AUC of 0.635 (SEM 0.041), surpassing other published methods in the AUC metric. Moreover, the proposed method yielded competitive results in other metrics. In external validation using the I-SPY 1 dataset, our approach achieved an AUC of 0.567 (SEM 0.032) for HR prediction and 0.563 (SEM 0.033) for HER2 prediction.

CONCLUSION

This study proposes a non-invasive system for identifying HER2 and HR in breast cancer. Although the results do not conclusively demonstrate superiority in both tasks, they indicate that the proposed method achieved good performance and is a competitive classifier compared to other reference methods. Ablation studies demonstrate that both radiomics features and deep features for the Siamese Neural Network are beneficial for the model. The introduced manufacturer bias normalization method has been shown to enhance the method's performance. Furthermore, the external validation of the method enhances the reliability of this research. Source code, pre-trained segmentation network, Radiomics and deep features, data for statistical analysis, and Supporting Information of this article are online at: https://github.com/FORRESTHUACHEN/Siamese_Neural_Network_based_Brest_cancer_Radiogenomic.

Unveiling the Links Between Microbial Alteration and Host Gene Disarray in Crohn's Disease via TAHMC

A compelling correlation method linking microbial communities and host gene expression in tissues is currently absent. A novel pipeline is proposed, dubbed Transcriptome Analysis of Host-Microbiome Crosstalk (TAHMC), designed to concurrently restore both host gene expression and microbial quantification from bulk RNA-seq data. Employing this approach, it discerned associations between the tissue microbiome and host immunity in the context of Crohn's disease (CD). Further, machine learning is utilized to separately construct networks of associations among host mRNA, long non-coding RNA, and tissue microbes. Unique host genes and tissue microbes are extracted from these networks for potential utility in CD diagnosis. Experimental validation of the predicted host gene regulation by microbes from the association network is achieved through the co-culturing of Faecalibacterium prausnitzii with Caco-2 cells. Collectively, the TAHMC pipeline accurately recovers both host gene expression and microbial quantification from CD RNA-seq data, thereby illuminating potential causal links between shifts in microbial composition as well as diversity within CD mucosal tissues and aberrant host gene expression.

Optimal batch determination for improved harmonization and prognostication of multi-center PET/CT radiomics feature in head and neck cancer

Objective. To determine the optimal approach for identifying and mitigating batch effects in PET/CT radiomics features, and further improve the prognosis of patients with head and neck cancer (HNC), this study investigated the performance of three batch harmonization methods.Approach. Unsupervised harmonization identified the batch labels by K-means clustering. Supervised harmonization regarding the image acquisition factors (center, manufacturer, scanner, filter kernel) as known/given batch labels, and Combat harmonization was then implemented separately and sequentially based on the batch labels, i.e. harmonizing features among batches determined by each factor individually or harmonizing features among batches determined by multiple factors successively. Extensive experiments were conducted to predict overall survival (OS) on public PET/CT datasets that contain 800 patients from 9 centers.Main results. In the external validation cohort, results show that compared to original models without harmonization, Combat harmonization would be beneficial in OS prediction with C-index of 0.687-0.740 versus 0.684-0.767. Supervised harmonization slightly outperformed unsupervised harmonization in all models (C-index: 0.692-0.767 versus 0.684-0.750). Separate harmonization outperformed sequential harmonization in CT_m+clinic and CT_cm+clinic models with C-index of 0.752 and 0.722, respectively, while sequential harmonization involved clinical features in PET_rs+clinic model further improving the performance and achieving the highest C-index of 0.767.Significance. Optimal batch determination especially sequential harmonization for Combat holds the potential to improve the prognostic power of radiomics model in multi-center HNC dataset with PET/CT imaging.

Development and External Validation of 18F-FDG PET-Based Radiomic Model for Predicting Pathologic Complete Response after Neoadjuvant Chemotherapy in Breast Cancer

The aim of our retrospective study is to develop and externally validate an 18F-FDG PET-derived radiomics model for predicting pathologic complete response (pCR) after neoadjuvant chemotherapy (NAC) in breast cancer patients. A total of 87 breast cancer patients underwent curative surgery after NAC at Soonchunhyang University Seoul Hospital and were randomly assigned to a training cohort and an internal validation cohort. Radiomic features were extracted from pretreatment PET images. A radiomic-score model was generated using the LASSO method. A combination model incorporating significant clinical variables was constructed. These models were externally validated in a separate cohort of 28 patients from Soonchunhyang University Buscheon Hospital. The model performances were assessed using area under the receiver operating characteristic (AUC). Seven radiomic features were selected to calculate the radiomic-score. Among clinical variables, human epidermal growth factor receptor 2 status was an independent predictor of pCR. The radiomic-score model achieved good discriminability, with AUCs of 0.963, 0.731, and 0.729 for the training, internal validation, and external validation cohorts, respectively. The combination model showed improved predictive performance compared to the radiomic-score model alone, with AUCs of 0.993, 0.772, and 0.906 in three cohorts, respectively. The 18F-FDG PET-derived radiomic-based model is useful for predicting pCR after NAC in breast cancer.

Harmonization of multi-site functional MRI data with dual-projection based ICA model

Modern neuroimaging studies frequently merge magnetic resonance imaging (MRI) data from multiple sites. A larger and more diverse group of participants can increase the statistical power, enhance the reliability and reproducibility of neuroimaging research, and obtain findings more representative of the general population. However, measurement biases caused by site differences in scanners represent a barrier when pooling data collected from different sites. The existence of site effects can mask biological effects and lead to spurious findings. We recently proposed a powerful denoising strategy that implements dual-projection (DP) theory based on ICA to remove site-related effects from pooled data, demonstrating the method for simulated and in vivo structural MRI data. This study investigates the use of our DP-based ICA denoising method for harmonizing functional MRI (fMRI) data collected from the Autism Brain Imaging Data Exchange II. After frequency-domain and regional homogeneity analyses, two modalities, including amplitude of low frequency fluctuation (ALFF) and regional homogeneity (ReHo), were used to validate our method. The results indicate that DP-based ICA denoising method removes unwanted site effects for both two fMRI modalities, with increases in the significance of the associations between non-imaging variables (age, sex, etc.) and fMRI measures. In conclusion, our DP method can be applied to fMRI data in multi-site studies, enabling more accurate and reliable neuroimaging research findings.

Single-cell and spatial transcriptomics: deciphering brain complexity in health and disease

In the past decade, single-cell technologies have proliferated and improved from their technically challenging beginnings to become common laboratory methods capable of determining the expression of thousands of genes in thousands of cells simultaneously. The field has progressed by taking the CNS as a primary research subject - the cellular complexity and multiplicity of neuronal cell types provide fertile ground for the increasing power of single-cell methods. Current single-cell RNA sequencing methods can quantify gene expression with sufficient accuracy to finely resolve even subtle differences between cell types and states, thus providing a great tool for studying the molecular and cellular repertoire of the CNS and its disorders. However, single-cell RNA sequencing requires the dissociation of tissue samples, which means that the interrelationships between cells are lost. Spatial transcriptomic methods bypass tissue dissociation and retain this spatial information, thereby allowing gene expression to be assessed across thousands of cells within the context of tissue structural organization. Here, we discuss how single-cell and spatially resolved transcriptomics have been contributing to unravelling the pathomechanisms underlying brain disorders. We focus on three areas where we feel these new technologies have provided particularly useful insights: selective neuronal vulnerability, neuroimmune dysfunction and cell-type-specific treatment response. We also discuss the limitations and future directions of single-cell and spatial RNA sequencing technologies.

Comprehensive Analysis to Reveal Amino Acid Metabolism-Associated Genes as a Prognostic Index in Gastric Cancer

Background

Amino acid metabolism (AAM) is related to tumor growth, prognosis, and therapeutic response. Tumor cells use more amino acids with less synthetic energy than normal cells for rapid proliferation. However, the possible significance of AAM-related genes in the tumor microenvironment (TME) is poorly understood.

Methods

Gastric cancer (GC) patients were classified into molecular subtypes by consensus clustering analysis using AAMs genes. AAM pattern, transcriptional patterns, prognosis, and TME in distinct molecular subtypes were systematically investigated. AAM gene score was built by least absolute shrinkage and selection operator (Lasso) regression.

Results

The study revealed that copy number variation (CNV) changes were prevalent in selected AAM-related genes, and most of these genes exhibited a high frequency of CNV deletion. Three molecular subtypes (clusters A, B, and C) were developed based on 99 AAM genes, which cluster B had better prognosis outcome. We developed a scoring system (AAM score) based on 4 AAM gene expressions to measure the AAM patterns of each patient. Importantly, we constructed a survival probability prediction nomogram. The AAM score was substantially associated with the index of cancer stem cells and sensitivity to chemotherapy intervention.

Conclusion

Overall, we detected prognostic AAM features in GC patients, which may help define TME characteristics and explore more effective treatment approaches.

Cross-scanner harmonization methods for structural MRI may need further work: A comparison study

The clinical usefulness MRI biomarkers for aging and dementia studies relies on precise brain morphological measurements; however, scanner and/or protocol variations may introduce noise or bias. One approach to address this is post-acquisition scan harmonization. In this work, we evaluate deep learning (neural style transfer, CycleGAN and CGAN), histogram matching, and statistical (ComBat and LongComBat) methods. Participants who had been scanned on both GE and Siemens scanners (cross-sectional participants, known as Crossover (n = 113), and longitudinally scanned participants on both scanners (n = 454)) were used. The goal was to match GE MPRAGE (T1-weighted) scans to Siemens improved resolution MPRAGE scans. Harmonization was performed on raw native and preprocessed (resampled, affine transformed to template space) scans. Cortical thicknesses were measured using FreeSurfer (v.7.1.1). Distributions were checked using Kolmogorov-Smirnov tests. Intra-class correlation (ICC) was used to assess the degree of agreement in the Crossover datasets and annualized percent change in cortical thickness was calculated to evaluate the Longitudinal datasets. Prior to harmonization, the least agreement was found at the frontal pole (ICC = 0.72) for the raw native scans, and at caudal anterior cingulate (0.76) and frontal pole (0.54) for the preprocessed scans. Harmonization with NST, CycleGAN, and HM improved the ICCs of the preprocessed scans at the caudal anterior cingulate (>0.81) and frontal poles (>0.67). In the Longitudinal raw native scans, over- and under-estimations of cortical thickness were observed due to the changing of the scanners. ComBat matched the cortical thickness distributions throughout but was not able to increase the ICCs or remove the effects of scanner changeover in the Longitudinal datasets. CycleGAN and NST performed slightly better to address the cortical thickness variations between scanner change. However, none of the methods succeeded in harmonizing the Longitudinal dataset. CGAN was the worst performer for both datasets. In conclusion, the performance of the methods was overall similar and region dependent. Future research is needed to improve the existing approaches since none of them outperformed each other in terms of harmonizing the datasets at all ROIs. The findings of this study establish framework for future research into the scan harmonization problem.

A multi-omic analysis reveals the esophageal dysbiosis as the predominant trait of eosinophilic esophagitis

BACKGROUND

Eosinophilic esophagitis (EoE) is a chronic immune-mediated rare disease, characterized by esophageal dysfunctions. It is likely to be primarily activated by food antigens and is classified as a chronic disease for most patients. Therefore, a deeper understanding of the pathogenetic mechanisms underlying EoE is needed to implement and improve therapeutic lines of intervention and ameliorate overall patient wellness.

METHODS

RNA-seq data of 18 different studies on EoE, downloaded from NCBI GEO with faster-qdump ( https://github.com/ncbi/sra-tools ), were batch-corrected and analyzed for transcriptomics and metatranscriptomics profiling as well as biological process functional enrichment. The EoE TaMMA web app was designed with plotly and dash. Tabula Sapiens raw data were downloaded from the UCSC Cell Browser. Esophageal single-cell raw data analysis was performed within the Automated Single-cell Analysis Pipeline. Single-cell data-driven bulk RNA-seq data deconvolution was performed with MuSiC and CIBERSORTx. Multi-omics integration was performed with MOFA.

RESULTS

The EoE TaMMA framework pointed out disease-specific molecular signatures, confirming its reliability in reanalyzing transcriptomic data, and providing new EoE-specific molecular markers including CXCL14, distinguishing EoE from gastroesophageal reflux disorder. EoE TaMMA also revealed microbiota dysbiosis as a predominant characteristic of EoE pathogenesis. Finally, the multi-omics analysis highlighted the presence of defined classes of microbial entities in subsets of patients that may participate in inducing the antigen-mediated response typical of EoE pathogenesis.

CONCLUSIONS

Our study showed that the complex EoE molecular network may be unraveled through advanced bioinformatics, integrating different components of the disease process into an omics-based network approach. This may implement EoE management and treatment in the coming years.

Transcriptomic Analyses of Pretreatment Tumor Biopsy Samples, Response to Neoadjuvant Chemoradiotherapy, and Survival in Patients With Advanced Rectal Cancer

Importance

Neoadjuvant chemoradiotherapy (CRT) is the standard of care for advanced rectal cancer. Yet, estimating response to CRT remains an unmet clinical challenge.

Objective

To investigate and better understand the transcriptomic factors associated with response to neoadjuvant CRT and survival in patients with advanced rectal cancer.

Design, Setting, and Participants

A single-center, retrospective, case series was conducted at a comprehensive cancer center. Pretreatment biopsies from 298 patients with rectal cancer who were later treated with neoadjuvant CRT between April 1, 2004, and September 30, 2020, were analyzed by RNA sequencing. Data analysis was performed from July 1, 2021, to May 31, 2022.

Exposures

Chemoradiotherapy followed by total mesorectal excision or watch-and-wait management.

Main Outcomes and Measures

Transcriptional subtyping was performed by consensus molecular subtype (CMS) classification. Immune cell infiltration was assessed using microenvironment cell populations-counter (MCP-counter) scores and single-sample gene set enrichment analysis (ssGSEA). Patients with surgical specimens of tumor regression grade 3 to 4 or whose care was managed by the watch-and-wait approach for more than 3 years were defined as good responders.

Results

Of the 298 patients in the study, 205 patients (68.8%) were men, and the median age was 61 (IQR, 52-67) years. Patients classified as CMS1 (6.4%) had a significantly higher rate of good response, albeit survival was comparable among the 4 subtypes. Good responders exhibited an enrichment in various immune-related pathways, as determined by ssGSEA. Microenvironment cell populations-counter scores for cytotoxic lymphocytes were significantly higher for good responders than nonresponders (median, 0.76 [IQR, 0.53-1.01] vs 0.58 [IQR, 0.43-0.83]; P < .001). Cytotoxic lymphocyte MCP-counter score was independently associated with response to CRT, as determined in the multivariable analysis (odds ratio, 3.81; 95% CI, 1.82-7.97; P < .001). Multivariable Cox proportional hazards regression analysis, including postoperative pathologic factors, revealed the cytotoxic lymphocyte MCP-counter score to be independently associated with recurrence-free survival (hazard ratio [HR], 0.38; 95% CI, 0.16-0.92; P = .03) and overall survival (HR, 0.16; 95% CI, 0.03-0.83; P = .03).

Conclusions and Relevance

In this case series of patients with rectal cancer treated with neoadjuvant CRT, the cytotoxic lymphocyte score in pretreatment biopsy samples, as computed by RNA sequencing, was associated with response to CRT and survival. This finding suggests that the cytotoxic lymphocyte score might serve as a biomarker in personalized multimodal rectal cancer treatment.

Normalization of Large-Scale Transcriptome Data Using Heuristic Methods

In this study, we introduce an artificial intelligent method for addressing the batch effect of a transcriptome data. The method has several clear advantages in comparison with the alternative methods presently in use. Batch effect refers to the discrepancy in gene expression data series, measured under different conditions. While the data from the same batch (measurements performed under the same conditions) are compatible, combining various batches into 1 data set is problematic because of incompatible measurements. Therefore, it is necessary to perform correction of the combined data (normalization), before performing biological analysis. There are numerous methods attempting to correct data set for batch effect. These methods rely on various assumptions regarding the distribution of the measurements. Forcing the data elements into pre-supposed distribution can severely distort biological signals, thus leading to incorrect results and conclusions. As the discrepancy between the assumptions regarding the data distribution and the actual distribution is wider, the biases introduced by such “correction methods” are greater. We introduce a heuristic method to reduce batch effect. The method does not rely on any assumptions regarding the distribution and the behavior of data elements. Hence, it does not introduce any new biases in the process of correcting the batch effect. It strictly maintains the integrity of measurements within the original batches.

Assessing the generalisability of radiomics features previously identified as predictive of radiation-induced sticky saliva and xerostomia

Background and purpose

While core to the scientific approach, reproducibility of experimental results is challenging in radiomics studies. A recent publication identified radiomics features that are predictive of late irradiation-induced toxicity in head and neck cancer (HNC) patients. In this study, we assessed the generalisability of these findings.

Materials and Methods

The procedure described in the publication in question was applied to a cohort of 109 HNC patients treated with 50-70 Gy in 20-35 fractions using helical radiotherapy although there were inherent differences between the two patient populations and methodologies. On each slice of the planning CT with delineated parotid and submandibular glands, the imaging features that were previously identified as predictive of moderate-to-severe xerostomia and sticky saliva 12 months post radiotherapy (Xer12m and SS12m) were calculated. Specifically, Short Run Emphasis (SRE) and maximum CT intensity (maxHU) were evaluated for improvement in prediction of Xer12m and SS12m respectively, compared to models solely using baseline toxicity and mean dose to the salivary glands.

Results

None of the associations previously identified as statistically significant and involving radiomics features in univariate or multivariate models could be reproduced on our cohort.

Conclusion

The discrepancies observed between the results of the two studies delineate limits to the generalisability of the previously reported findings. This may be explained by the differences in the approaches, in particular the imaging characteristics and subsequent methodological implementation. This highlights the importance of external validation, high quality reporting guidelines and standardisation protocols to ensure generalisability, replication and ultimately clinical implementation.

Site effects how-to and when: An overview of retrospective techniques to accommodate site effects in multi-site neuroimaging analyses

Site differences, or systematic differences in feature distributions across multiple data-acquisition sites, are a known source of heterogeneity that may adversely affect large-scale meta- and mega-analyses of independently collected neuroimaging data. They influence nearly all multi-site imaging modalities and biomarkers, and methods to compensate for them can improve reliability and generalizability in the analysis of genetics, omics, and clinical data. The origins of statistical site effects are complex and involve both technical differences (scanner vendor, head coil, acquisition parameters, imaging processing) and differences in sample characteristics (inclusion/exclusion criteria, sample size, ancestry) between sites. In an age of expanding international consortium research, there is a growing need to disentangle technical site effects from sample characteristics of interest. Numerous statistical and machine learning methods have been developed to control for, model, or attenuate site effects - yet to date, no comprehensive review has discussed the benefits and drawbacks of each for different use cases. Here, we provide an overview of the different existing statistical and machine learning methods developed to remove unwanted site effects from independently collected neuroimaging samples. We focus on linear mixed effect models, the ComBat technique and its variants, adjustments based on image quality metrics, normative modeling, and deep learning approaches such as generative adversarial networks. For each method, we outline the statistical foundation and summarize strengths and weaknesses, including their assumptions and conditions of use. We provide information on software availability and comment on the ease of use and the applicability of these methods to different types of data. We discuss validation and comparative reports, mention caveats and provide guidance on when to use each method, depending on context and specific research questions.

The spatio-temporal evolution of multiple myeloma from baseline to relapse-refractory states

Deciphering Multiple Myeloma evolution in the whole bone marrow is key to inform curative strategies. Here, we perform spatial-longitudinal whole-exome sequencing, including 140 samples collected from 24 Multiple Myeloma patients during up to 14 years. Applying imaging-guided sampling we observe three evolutionary patterns, including relapse driven by a single-cell expansion, competing/co-existing sub-clones, and unique sub-clones at distinct locations. While we do not find the unique relapse sub-clone in the baseline focal lesion(s), we show a close phylogenetic relationship between baseline focal lesions and relapse disease, highlighting focal lesions as hotspots of tumor evolution. In patients with ≥3 focal lesions on positron-emission-tomography at diagnosis, relapse is driven by multiple distinct sub-clones, whereas in other patients, a single-cell expansion is typically seen (p < 0.01). Notably, we observe resistant sub-clones that can be hidden over years, suggesting that a prerequisite for curative therapies would be to overcome not only tumor heterogeneity but also dormancy.

AutoComBat: a generic method for harmonizing MRI-based radiomic features

The use of multicentric data is becoming essential for developing generalizable radiomic signatures. In particular, Magnetic Resonance Imaging (MRI) data used in brain oncology are often heterogeneous in terms of scanners and acquisitions, which significantly impact quantitative radiomic features. Various methods have been proposed to decrease dependency, including methods acting directly on MR images, i.e., based on the application of several preprocessing steps before feature extraction or the ComBat method, which harmonizes radiomic features themselves. The ComBat method used for radiomics may be misleading and presents some limitations, such as the need to know the labels associated with the "batch effect". In addition, a statistically representative sample is required and the applicability of a signature whose batch label is not present in the train set is not possible. This work aimed to compare a priori and a posteriori radiomic harmonization methods and propose a code adaptation to be machine learning compatible. Furthermore, we have developed AutoComBat, which aims to automatically determine the batch labels, using either MRI metadata or quality metrics as inputs of the proposed constrained clustering. A heterogeneous dataset consisting of high and low-grade gliomas coming from eight different centers was considered. The different methods were compared based on their ability to decrease relative standard deviation of radiomic features extracted from white matter and on their performance on a classification task using different machine learning models. ComBat and AutoComBat using image-derived quality metrics as inputs for batch assignment and preprocessing methods presented promising results on white matter harmonization, but with no clear consensus for all MR images. Preprocessing showed the best results on the T1w-gd images for the grading task. For T2w-flair, AutoComBat, using either metadata plus quality metrics or metadata alone as inputs, performs better than the conventional ComBat, highlighting its potential for data harmonization. Our results are MRI weighting, feature class and task dependent and require further investigations on other datasets.

Identification of a Five-mRNA Signature as a Novel Potential Prognostic Biomarker for Glioblastoma by Integrative Analysis

Despite the availability of advanced multimodal therapy, the prognosis of patients suffering from glioblastoma (GBM) remains poor. We conducted a genome-wide integrative analysis of mRNA expression profiles in 302 GBM tissues and 209 normal brain tissues from the Gene Expression Omnibus (GEO), The Cancer Genome Atlas (TCGA), and the Genotype-Tissue Expression (GTEx) project to examine the prognostic and predictive value of specific mRNAs in GBM. A total of 26 mRNAs were identified to be closely related to GBM patients’ OS (p < 0.05). Utilizing survival analysis and the Cox regression model, we discovered a set of five mRNAs (PTPRN, ABCC3, MDK, NMB, and RALYL) from these 26 mRNAs that displayed the capacity to stratify patients into high- and low-risk groups with statistically different overall survival in the training set. The model of the five-mRNA biomarker signature was successfully verified on a testing set and independent sets. Moreover, multivariate Cox regression analysis revealed that the five-mRNA biomarker signature was a prognostic factor for the survival of patients with GBM independent of clinical characteristics and molecular features (p < 0.05). Gene set enrichment analysis indicated that the five-mRNA biomarker signature might be implicated in the incidence and development of GBM through its roles in known cancer-related pathways, signaling molecules, and the immune system. Moreover, consistent with the bioinformatics analysis, NMB, ABCC3, and MDK mRNA expression was considerably higher in four human GBM cells, and the expression of PTPRN and RALYL was decreased in GBM cells (p < 0.05). Our study developed a novel candidate model that provides new prospective prognostic biomarkers for GBM.

Data harmonisation for information fusion in digital healthcare: A state-of-the-art systematic review, meta-analysis and future research directions

Removing the bias and variance of multicentre data has always been a challenge in large scale digital healthcare studies, which requires the ability to integrate clinical features extracted from data acquired by different scanners and protocols to improve stability and robustness. Previous studies have described various computational approaches to fuse single modality multicentre datasets. However, these surveys rarely focused on evaluation metrics and lacked a checklist for computational data harmonisation studies. In this systematic review, we summarise the computational data harmonisation approaches for multi-modality data in the digital healthcare field, including harmonisation strategies and evaluation metrics based on different theories. In addition, a comprehensive checklist that summarises common practices for data harmonisation studies is proposed to guide researchers to report their research findings more effectively. Last but not least, flowcharts presenting possible ways for methodology and metric selection are proposed and the limitations of different methods have been surveyed for future research.

Dysregulation of Human Somatic piRNA Expression in Parkinson's Disease Subtypes and Stages

Piwi interacting RNAs (piRNAs) are small non-coding single-stranded RNA species 20–31 nucleotides in size generated from distinct loci. In germline tissues, piRNAs are amplified via a “ping-pong cycle” to produce secondary piRNAs, which act in transposon silencing. In contrast, the role of somatic-derived piRNAs remains obscure. Here, we investigated the identity and distribution of piRNAs in human somatic tissues to determine their function and potential role in Parkinson’s disease (PD). Human datasets were curated from the Gene Expression Omnibus (GEO) database and a workflow was developed to identify piRNAs, which revealed 902 somatic piRNAs of which 527 were expressed in the brain. These were mainly derived from chromosomes 1, 11, and 19 compared to the germline tissues, which were from 15 and 19. Approximately 20% of somatic piRNAs mapped to transposon 3′ untranslated regions (UTRs), but a large proportion were sensed to the transcript in contrast to germline piRNAs. Gene set enrichment analysis suggested that somatic piRNAs function in neurodegenerative disease. piRNAs undergo dysregulation in different PD subtypes (PD and Parkinson’s disease dementia (PDD)) and stages (premotor and motor). piR-has-92056, piR-hsa-150797, piR-hsa-347751, piR-hsa-1909905, piR-hsa-2476630, and piR-hsa-2834636 from blood small extracellular vesicles were identified as novel biomarkers for PD diagnosis using a sparse partial least square discriminant analysis (sPLS-DA) (accuracy: 92%, AUC = 0.89). This study highlights a role for piRNAs in PD and provides tools for novel biomarker development.

18F-FDOPA PET for the Noninvasive Prediction of Glioma Molecular Parameters: A Radiomics Study

The assessment of gliomas by 18F-FDOPA PET imaging as an adjunct to MRI showed high performance by combining static and dynamic features to noninvasively predict the isocitrate dehydrogenase (IDH) mutations and the 1p/19q codeletion, which the World Health Organization classified as significant parameters in 2016. The current study evaluated whether other 18F-FDOPA PET radiomics features further improve performance and the contributions of each of these features to performance. Methods: Our study included 72 retrospectively selected, newly diagnosed glioma patients with 18F-FDOPA PET dynamic acquisitions. A set of 114 features, including conventional static features and dynamic features, as well as other radiomics features, were extracted and machine-learning models trained to predict IDH mutations and the 1p/19q codeletion. Models were based on a machine-learning algorithm built from stable, relevant, and uncorrelated features selected by hierarchic clustering followed by a bootstrapped feature selection process. Models were assessed by comparing area under the curve using a nested cross-validation approach. Feature importance was assessed using Shapley additive explanations values. Results: The best models were able to predict IDH mutations (logistic regression with L2 regularization) and the 1p/19q codeletion (support vector machine with radial basis function kernel) with an area under the curve of 0.831 (95% CI, 0.790-0.873) and 0.724 (95% CI, 0.669-0.782), respectively. For the prediction of IDH mutations, dynamic features were the most important features in the model (time to peak, 35.5%). In contrast, other radiomics features were the most useful for predicting the 1p/19q codeletion (up to 14.5% of importance for the small-zone low-gray-level emphasis). Conclusion:18F-FDOPA PET is an effective tool for the noninvasive prediction of glioma molecular parameters using a full set of amino-acid PET radiomics features. The contribution of each feature set shows the importance of systematically integrating dynamic acquisition for prediction of the IDH mutations as well as developing the use of radiomics features in routine practice for prediction of the 1p/19q codeletion.

Multi-Stage Harmonization for Robust AI across Breast MR Databases

Simple Summary

Batch harmonization of radiomic features extracted from magnetic resonance images of breast lesions from two databases was applied to an artificial intelligence/machine learning classification workflow. Training and independent test sets from the two databases, as well as the combination of them, were used in pre-harmonization and post-harmonization forms to investigate the generalizability of performance in the task of distinguishing between malignant and benign lesions. Most training and independent test scenarios were statistically equivalent, demonstrating that batch harmonization with feature selection harmonization can potentially develop generalizable classification models.

Abstract

Radiomic features extracted from medical images may demonstrate a batch effect when cases come from different sources. We investigated classification performance using training and independent test sets drawn from two sources using both pre-harmonization and post-harmonization features. In this retrospective study, a database of thirty-two radiomic features, extracted from DCE-MR images of breast lesions after fuzzy c-means segmentation, was collected. There were 944 unique lesions in Database A (208 benign lesions, 736 cancers) and 1986 unique lesions in Database B (481 benign lesions, 1505 cancers). The lesions from each database were divided by year of image acquisition into training and independent test sets, separately by database and in combination. ComBat batch harmonization was conducted on the combined training set to minimize the batch effect on eligible features by database. The empirical Bayes estimates from the feature harmonization were applied to the eligible features of the combined independent test set. The training sets (A, B, and combined) were then used in training linear discriminant analysis classifiers after stepwise feature selection. The classifiers were then run on the A, B, and combined independent test sets. Classification performance was compared using pre-harmonization features to post-harmonization features, including their corresponding feature selection, evaluated using the area under the receiver operating characteristic curve (AUC) as the figure of merit. Four out of five training and independent test scenarios demonstrated statistically equivalent classification performance when compared pre- and post-harmonization. These results demonstrate that translation of machine learning techniques with batch data harmonization can potentially yield generalizable models that maintain classification performance.

A guide to ComBat harmonization of imaging biomarkers in multicenter studies

The impact of PET image acquisition and reconstruction parameters on SUV measurements or radiomic feature values is widely documented. This "scanner" effect is detrimental to the design and validation of predictive or prognostic models and limits the use of large multicenter cohorts. To reduce the impact of this scanner effect, the ComBat method has been proposed and is now used in various contexts. The purpose of this article is to explain and illustrate the use of ComBat based on practical examples. We also give examples in which the ComBat assumptions are not met; thus, ComBat should not be used.

The Inflammatory Bowel Disease Transcriptome and Metatranscriptome Meta-Analysis (IBD TaMMA) framework

Inflammatory bowel disease (IBD) is a class of chronic disorders whose etiogenesis is still unknown. Despite the high number of IBD-related omics studies, the RNA-sequencing data produced results that are hard to compare because of the experimental variability and different data analysis approaches. We here introduce the IBD Transcriptome and Metatranscriptome Meta-Analysis (TaMMA) framework, a comprehensive survey of publicly available IBD RNA-sequencing datasets. IBD TaMMA is an open-source platform where scientists can explore simultaneously the freely available IBD-associated transcriptomics and microbial profiles thanks to its interactive interface, resulting in a useful tool to the IBD community.

On the feasibility of deep learning applications using raw mass spectrometry data

SUMMARY

In recent years, SWATH-MS has become the proteomic method of choice for data-independent-acquisition, as it enables high proteome coverage, accuracy and reproducibility. However, data analysis is convoluted and requires prior information and expert curation. Furthermore, as quantification is limited to a small set of peptides, potentially important biological information may be discarded. Here we demonstrate that deep learning can be used to learn discriminative features directly from raw MS data, eliminating hence the need of elaborate data processing pipelines. Using transfer learning to overcome sample sparsity, we exploit a collection of publicly available deep learning models already trained for the task of natural image classification. These models are used to produce feature vectors from each mass spectrometry (MS) raw image, which are later used as input for a classifier trained to distinguish tumor from normal prostate biopsies. Although the deep learning models were originally trained for a completely different classification task and no additional fine-tuning is performed on them, we achieve a highly remarkable classification performance of 0.876 AUC. We investigate different types of image preprocessing and encoding. We also investigate whether the inclusion of the secondary MS2 spectra improves the classification performance. Throughout all tested models, we use standard protein expression vectors as gold standards. Even with our naïve implementation, our results suggest that the application of deep learning and transfer learning techniques might pave the way to the broader usage of raw mass spectrometry data in real-time diagnosis.

AVAILABILITY AND IMPLEMENTATION

The open source code used to generate the results from MS images is available on GitHub: https://ibm.biz/mstransc. The raw MS data underlying this article cannot be shared publicly for the privacy of individuals that participated in the study. Processed data including the MS images, their encodings, classification labels and results can be accessed at the following link: https://ibm.box.com/v/mstc-supplementary.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

A transfer learning approach to facilitate ComBat-based harmonization of multicentre radiomic features in new datasets

PURPOSE

To facilitate the demonstration of the prognostic value of radiomics, multicenter radiomics studies are needed. Pooling radiomic features of such data in a statistical analysis is however challenging, as they are sensitive to the variability in scanner models, acquisition protocols and reconstruction settings, which is often unavoidable in a multicentre retrospective analysis. A statistical harmonization strategy called ComBat was utilized in radiomics studies to deal with the "center-effect". The goal of the present work was to integrate a transfer learning (TL) technique within ComBat-and recently developed alternate versions of ComBat with improved flexibility (M-ComBat) and robustness (B-ComBat)-to allow the use of a previously determined harmonization transform to the radiomic feature values of new patients from an already known center.

MATERIAL AND METHODS

The proposed TL approach were incorporated in the four versions of ComBat (standard, B, M, and B-M ComBat). The proposed approach was evaluated using a dataset of 189 locally advanced cervical cancer patients from 3 centers, with magnetic resonance imaging (MRI) and positron emission tomography (PET) images, with the clinical endpoint of predicting local failure. The impact performance of the TL approach was evaluated by comparing the harmonization achieved using only parts of the data to the reference (harmonization achieved using all the available data). It was performed through three different machine learning pipelines.

RESULTS

The proposed TL technique was successful in harmonizing features of new patients from a known center in all versions of ComBat, leading to predictive models reaching similar performance as the ones developed using the features harmonized with all the data available.

CONCLUSION

The proposed TL approach enables applying a previously determined ComBat transform to new, previously unseen data.

Evaluation of immune status in testis and macrophage polarization associated with testicular damage in patients with nonobstructive azoospermia

OBJECTIVE

Immune cells residing in the testicular interstitial space form the immunological microenvironment of the testis. They are assumed to play a role in maintaining testicular homeostasis and immune privilege. However, the immune status and related cell polarization in patients with nonobstructive azoospermia (NOA) remains poorly characterized. System evaluation of the testis immunological microenvironment in NOA patients may help to reveal the mechanisms of idiopathic azoospermia.

STUDY DESIGN

The gene expression patterns of immune cells in normal human testes were systematically analyzed by single-cell RNA sequencing (scRNA-seq) and preliminarily verification by the human protein atlas (HPA) online database. The immune cell infiltration profiles and immune status of patients with NOA was analyzed by single-sample gene set enrichment analysis (ssGSEA) and gene set variation analysis (GSVA) based on four independent public microarray datasets (GSE45885, GSE45887, GSE9210, and GSE145467), obtained from Gene Expression Omnibus (GEO) online database. The relationship between immune cells and spermatogenesis score was further analyzed by Spearman correlation analysis. Finally, immunohistochemistry (IHC) staining was performed to identify the main immune cell types and their polarization status in patients with NOA.

RESULTS

Both scRNA-seq and HPA analysis showed that testicular macrophages represent the largest pool of immune cells in the normal testis, and also exhibit an attenuated inflammatory response by expressing high levels of tolerance proteins (CD163, IL-10, TGF-β, and VEGF) and reduced expression of TLR signaling pathway-related genes. Correlation analysis revealed that the testicular immune score and macrophages including M1 and M2 macrophages were significantly negatively correlated with spermatogenesis score in patients with NOA (GSE45885 and GSE45887). In addition, the number of M1 and M2 macrophages was significantly higher in patients with NOA (GSE9210 and GSE145467) than in normal testis. GSVA analysis indicated that the immunological microenvironment in NOA tissues was manifested by activated immune system and pro-inflammatory status. IHC staining results showed that the number of M1 and M2 macrophages was significantly higher in NOA tissues than in normal testis and negatively correlated with the Johnson score.

CONCLUSION

Testicular macrophage polarization may play a vital role in NOA development and is a promising potential therapeutic target.

Tranexamic acid: current use in obstetrics, major orthopedic, and trauma surgery

PURPOSE

In this Continuing Professional Development module, we review the practical pharmacology of tranexamic acid and its clinical use in trauma, obstetrics, and major orthopedic surgery.

PRINCIPAL FINDINGS

Tranexamic acid is a synthetic drug that inhibits fibrinolysis. Multiple clinical trials in various clinical settings have shown that it can reduce blood loss, transfusion rates, and bleeding-associated mortality. In trauma and obstetrical bleeding, early tranexamic acid administration (< three hours) may have greater clinical benefits. Overall, tranexamic acid use appears safe with no significant increase of thromboembolic or seizure events. Nevertheless, current evidence has limitations related to wide heterogeneity in dose, route, and timing of drug administration, as well as generalizability of the large-scale trial findings to higher income nations.

CONCLUSIONS

Tranexamic acid is an efficacious and safe pharmacological-based blood conservation technique in the management of clinically significant hemorrhage. All anesthesiologists should have a good understanding of the pharmacotherapeutic properties and perioperative role of tranexamic acid therapy both inside and outside of the operating room. The use of tranexamic acid is likely to continue to rise with endorsement by various clinical guidelines and healthcare organizations. Further quantitative research is needed to evaluate optimal dosing and drug efficacy in these clinical scenarios.

DBnorm as an R package for the comparison and selection of appropriate statistical methods for batch effect correction in metabolomic studies

As a powerful phenotyping technology, metabolomics provides new opportunities in biomarker discovery through metabolome-wide association studies (MWAS) and the identification of metabolites having a regulatory effect in various biological processes. While mass spectrometry-based (MS) metabolomics assays are endowed with high throughput and sensitivity, MWAS are doomed to long-term data acquisition generating an overtime-analytical signal drift that can hinder the uncovering of real biologically relevant changes. We developed “dbnorm”, a package in the R environment, which allows for an easy comparison of the model performance of advanced statistical tools commonly used in metabolomics to remove batch effects from large metabolomics datasets. “dbnorm” integrates advanced statistical tools to inspect the dataset structure not only at the macroscopic (sample batches) scale, but also at the microscopic (metabolic features) level. To compare the model performance on data correction, “dbnorm” assigns a score that help users identify the best fitting model for each dataset. In this study, we applied “dbnorm” to two large-scale metabolomics datasets as a proof of concept. We demonstrate that “dbnorm” allows for the accurate selection of the most appropriate statistical tool to efficiently remove the overtime signal drift and to focus on the relevant biological components of complex datasets.

Multi-omics characterization of WNT pathway reactivation to ameliorate BET inhibitor resistance in liver cancer cells

The Bromodomain and Extra-terminal domain (BET) proteins are promising targets in treating cancers. Although BET inhibitors have been in clinical trials, they are limited by lacking of suitable biomarkers to indicate drug responses in different cancers. Here we identify DHRS2, ETV4 and NOTUM as potential biomarkers to indicate drug resistance in liver cancer cells of a recently discovered BET inhibitor, Hjp-6-171. Furthermore, we confirm that reactivation of WNT pathway, the target of NOTUM, contributes to the drug sensitivity restoration in Hjp-6-171 resistant cells. Specially, combinations of Hjp-6-171 and a GSK3β inhibitor CHIR-98014 show remarkable therapeutic effects in vitro and in vivo. Integrating RNA-seq and ChIP-seq data, we reveal the expression signature of β-catenin regulated genes is contrary in sensitive cells to that in resistant cells. We propose WNT signaling molecules such as β-catenin and ETV4 to be candidate biomarkers to indicate BET inhibitor responses in liver cancer patients.

Tutorial: guidelines for the computational analysis of single-cell RNA sequencing data

Single-cell RNA sequencing (scRNA-seq) is a popular and powerful technology that allows you to profile the whole transcriptome of a large number of individual cells. However, the analysis of the large volumes of data generated from these experiments requires specialized statistical and computational methods. Here we present an overview of the computational workflow involved in processing scRNA-seq data. We discuss some of the most common tasks and the tools available for addressing central biological questions. In this article and our companion website ( https://scrnaseq-course.cog.sanger.ac.uk/website/index.html ), we provide guidelines regarding best practices for performing computational analyses. This tutorial provides a hands-on guide for experimentalists interested in analyzing their data as well as an overview for bioinformaticians seeking to develop new computational methods.

Development of a prognostic signature for bladder cancer based on immune-related genes

Background

Although the prognosis of patients with bladder cancer (BC) has improved significantly with the use of multimodal therapy, reliable prognostic biomarkers are still urgently needed due to the heterogeneity of tumors. Our aim was to develop an individualized immune-related gene pair (IRGP) signature that could precisely predict prognosis in BC patients.

Methods

Gene expression profiles and corresponding clinical information were collected from eight microarray data sets and one RNA-Seq data set.

Results

Among 1,811 immune genes, a 30-IRGP signature consisting of 52 unique genes was generated in the training cohort, which significantly stratified patients into low- and high-risk groups in terms of overall survival. In the testing and validation cohorts, the IRGP signature was also associated with patient prognosis in the univariate and multivariate Cox regression analyses. Several biological processes, including the immune response, chemotaxis, and the inflammatory response, were enriched among genes in the IRGP signature. When the signature was integrated with the TNM stage, an IRGP nomogram was developed and showed improved prognostic accuracy relative to the IRGP signature alone.

Conclusions

In short, we identified a robust IRGP signature for estimating overall survival in BC patients that could also be used as a promising biomarker for identifying high-risk patients for individualized therapy.

Random survival forest to predict transplant-eligible newly diagnosed multiple myeloma outcome including FDG-PET radiomics: a combined analysis of two independent prospective European trials

PURPOSE

Fluorodeoxyglucose-positron emission tomography/computed tomography (FDG-PET/CT) is included in the International Myeloma Working Group (IMWG) imaging guidelines for the work-up at diagnosis and the follow-up of multiple myeloma (MM) notably because it is a reliable tool as a predictor of prognosis. Nevertheless, none of the published studies focusing on the prognostic value of PET-derived features at baseline consider tumor heterogeneity, which could be of high importance in MM. The aim of this study was to evaluate the prognostic value of baseline PET-derived features in transplant-eligible newly diagnosed (TEND) MM patients enrolled in two prospective independent European randomized phase III trials using an innovative statistical random survival forest (RSF) approach.

METHODS

Imaging ancillary studies of IFM/DFCI2009 and EMN02/HO95 trials formed part of the present analysis (IMAJEM and EMN02/HO95, respectively). Among all patients initially enrolled in these studies, those with a positive baseline FDG-PET/CT imaging and focal bone lesions (FLs) and/or extramedullary disease (EMD) were included in the present analysis. A total of 17 image features (visual and quantitative, reflecting whole imaging characteristics) and 5 clinical/histopathological parameters were collected. The statistical analysis was conducted using two RSF approaches (train/validation + test and additional nested cross-validation) to predict progression-free survival (PFS).

RESULTS

One hundred thirty-nine patients were considered for this study. The final model based on the first RSF (train/validation + test) approach selected 3 features (treatment arm, hemoglobin, and SUV_maxBone Marrow (BM)) among the 22 involved initially, and two risk groups of patients (good and poor prognosis) could be defined with a mean hazard ratio of 4.3 ± 1.5 and a mean log-rank p value of 0.01 ± 0.01. The additional RSF (nested cross-validation) analysis highlighted the robustness of the proposed model across different splits of the dataset. Indeed, the first features selected using the train/validation + test approach remained the first ones over the folds with the nested approach.

CONCLUSION

We proposed a new prognosis model for TEND MM patients at diagnosis based on two RSF approaches.

TRIAL REGISTRATION

IMAJEM: NCT01309334 and EMN02/HO95: NCT01134484.

Glucose Metabolism Quantified by SUVmax on Baseline FDG-PET/CT Predicts Survival in Newly Diagnosed Multiple Myeloma Patients: Combined Harmonized Analysis of Two Prospective Phase III Trials

Simple Summary

Multiple myeloma (MM) is associated with high morbidity and mortality and variable survival that requires early identification of high-risk patients in order to quickly adapt treatment. FDG-PET/CT is a promising technique for initial staging of symptomatic MM. The aim of our retrospective study was to asses the prognostic value of this technique at baseline in symptomatic MM patients included in two large European prospective studies. After harmonization of data by and ad-hoc approach called M-Combat, we confirmed the prognostic value of FDG-PET/CT in a population of 227 MM patients, by integrating a new prognostic biomarker named “bone SUVmax” (including the maximum intensity of fixation of focal lesions and bone marrow) which is strongly correlated with a poorer prognosis of MM patients. Prognostic patient stratification is currently based on laboratory tests and genomic abnormalities, but FDG-PET/CT is likely to be an important method of defining high-risk patients, and thus, to potentially better adapt future therapeutic management.

Abstract

Background: Multiple myeloma is a hematological neoplasm characterized by a clonal proliferation of malignant plasma cells in the bone marrow, and is associated with high morbidity and mortality and variable survival. Positron emission tomography combined with computed tomography using 18F-deoxyfluoroglucose (FDG-PET/CT) is a promising technique for initial staging of symptomatic multiple myeloma patients. The objective of this study was to assess the prognostic value of this technique at baseline in symptomatic multiple myeloma patients included in two large European prospective studies (French and Italian). Methods: We retrospectively performed a combined harmonized analysis of 227 newly diagnosed transplant eligible multiple myeloma patients from two separate phase III trials. All images were centrally reviewed and analyzed using visual criteria and maximal standardized uptake value. An ad-hoc approach (called modified Combat) was applied to harmonize the data and then remove the “country effect” in order to strengthen the reliability of the final conclusions. Results: Using a multivariate analysis including treatment arm, R-ISS score, presence of extra-medullary disease and bone SUVmax, only bone SUVmax (p = 0.016) was an independent prognosis factor with an OS threshold of 7.1. For PFS, treatment arm and presence of extra-medullary disease were both independent prognosis biomarkers (p = 0.022 and 0.006 respectively). Conclusions: Our results show that bone SUVmax is a simple and reliable biomarker to analyze FDG-PET/CT at baseline that strongly correlates with a poorer prognosis for MM patients.

Colorectal Cancer Prediction Based on Weighted Gene Co-Expression Network Analysis and Variational Auto-Encoder

An effective feature extraction method is key to improving the accuracy of a prediction model. From the Gene Expression Omnibus (GEO) database, which includes 13,487 genes, we obtained microarray gene expression data for 238 samples from colorectal cancer (CRC) samples and normal samples. Twelve gene modules were obtained by weighted gene co-expression network analysis (WGCNA) on 173 samples. By calculating the Pearson correlation coefficient (PCC) between the characteristic genes of each module and colorectal cancer, we obtained a key module that was highly correlated with CRC. We screened hub genes from the key module by considering module membership, gene significance, and intramodular connectivity. We selected 10 hub genes as a type of feature for the classifier. We used the variational autoencoder (VAE) for 1159 genes with significantly different expressions and mapped the data into a 10-dimensional representation, as another type of feature for the cancer classifier. The two types of features were applied to the support vector machines (SVM) classifier for CRC. The accuracy was 0.9692 with an AUC of 0.9981. The result shows a high accuracy of the two-step feature extraction method, which includes obtaining hub genes by WGCNA and a 10-dimensional representation by variational autoencoder (VAE).

GDASC: a GPU parallel-based web server for detecting hidden batch factors

SUMMARY

We developed GDASC, a web version of our former DASC algorithm implemented with GPU. It provides a user-friendly web interface for detecting batch factors. Based on the good performance of DASC algorithm, it is able to give the most accurate results. For two steps of DASC, data-adaptive shrinkage and semi-non-negative matrix factorization, we designed parallelization strategies facing convex clustering solution and decomposition process. It runs more than 50 times faster than the original version on the representative RNA sequencing quality control dataset. With its accuracy and high speed, this server will be a useful tool for batch effects analysis.

AVAILABILITY AND IMPLEMENTATION

http://bioinfo.nankai.edu.cn/gdasc.php.

CONTACT

zhanghan@nankai.edu.cn.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

Therapeutic Angiogenesis with Sound Waves

Along with the progress of global aging, the prognosis of severe ischemic heart disease (IHD) remains poor, and thus the development of effective angiogenic therapy remains an important clinical unmet need. We have developed low-energy extracorporeal cardiac shock wave therapy as an innovative minimally invasive angiogenic therapy and confirmed its efficacy in a porcine chronic myocardial ischemia model in animal experiments as well as in patients with refractory angina. Since ultrasound is more advantageous for clinical application than shock waves, we then aimed to develop ultrasound therapy for IHD. We demonstrated that specific conditions of low-intensity pulsed ultrasound (LIPUS) therapy improve myocardial ischemia in animal models through the enhancement of angiogenesis mediated by endothelial mechanotransduction. To examine the effectiveness of our LIPUS therapy in patients with severe angina pectoris, we are now conducting a prospective multicenter clinical trial in Japan. Furthermore, to overcome the current serious situation of dementia pandemic but with no effective treatments worldwide, we have recently demonstrated that our LIPUS therapy also improves cognitive impairment in mouse models of Alzheimer's disease and vascular dementia. Here, we summarize the progress in our studies to develop angiogenic therapies with sound waves.

Performance comparison of modified ComBat for harmonization of radiomic features for multicenter studies

Multicenter studies are needed to demonstrate the clinical potential value of radiomics as a prognostic tool. However, variability in scanner models, acquisition protocols and reconstruction settings are unavoidable and radiomic features are notoriously sensitive to these factors, which hinders pooling them in a statistical analysis. A statistical harmonization method called ComBat was developed to deal with the "batch effect" in gene expression microarray data and was used in radiomics studies to deal with the "center-effect". Our goal was to evaluate modifications in ComBat allowing for more flexibility in choosing a reference and improving robustness of the estimation. Two modified ComBat versions were evaluated: M-ComBat allows to transform all features distributions to a chosen reference, instead of the overall mean, providing more flexibility. B-ComBat adds bootstrap and Monte Carlo for improved robustness in the estimation. BM-ComBat combines both modifications. The four versions were compared regarding their ability to harmonize features in a multicenter context in two different clinical datasets. The first contains 119 locally advanced cervical cancer patients from 3 centers, with magnetic resonance imaging and positron emission tomography imaging. In that case ComBat was applied with 3 labels corresponding to each center. The second one contains 98 locally advanced laryngeal cancer patients from 5 centers with contrast-enhanced computed tomography. In that specific case, because imaging settings were highly heterogeneous even within each of the five centers, unsupervised clustering was used to determine two labels for applying ComBat. The impact of each harmonization was evaluated through three different machine learning pipelines for the modelling step in predicting the clinical outcomes, across two performance metrics (balanced accuracy and Matthews correlation coefficient). Before harmonization, almost all radiomic features had significantly different distributions between labels. These differences were successfully removed with all ComBat versions. The predictive ability of the radiomic models was always improved with harmonization and the improved ComBat provided the best results. This was observed consistently in both datasets, through all machine learning pipelines and performance metrics. The proposed modifications allow for more flexibility and robustness in the estimation. They also slightly but consistently improve the predictive power of resulting radiomic models.

ProteomicsDB: a multi-omics and multi-organism resource for life science research

ProteomicsDB (https://www.ProteomicsDB.org) started as a protein-centric in-memory database for the exploration of large collections of quantitative mass spectrometry-based proteomics data. The data types and contents grew over time to include RNA-Seq expression data, drug-target interactions and cell line viability data. In this manuscript, we summarize new developments since the previous update that was published in Nucleic Acids Research in 2017. Over the past two years, we have enriched the data content by additional datasets and extended the platform to support protein turnover data. Another important new addition is that ProteomicsDB now supports the storage and visualization of data collected from other organisms, exemplified by Arabidopsis thaliana. Due to the generic design of ProteomicsDB, all analytical features available for the original human resource seamlessly transfer to other organisms. Furthermore, we introduce a new service in ProteomicsDB which allows users to upload their own expression datasets and analyze them alongside with data stored in ProteomicsDB. Initially, users will be able to make use of this feature in the interactive heat map functionality as well as the drug sensitivity prediction, but ultimately will be able to use all analytical features of ProteomicsDB in this way.

STATegra, a comprehensive multi-omics dataset of B-cell differentiation in mouse

Multi-omics approaches use a diversity of high-throughput technologies to profile the different molecular layers of living cells. Ideally, the integration of this information should result in comprehensive systems models of cellular physiology and regulation. However, most multi-omics projects still include a limited number of molecular assays and there have been very few multi-omic studies that evaluate dynamic processes such as cellular growth, development and adaptation. Hence, we lack formal analysis methods and comprehensive multi-omics datasets that can be leveraged to develop true multi-layered models for dynamic cellular systems. Here we present the STATegra multi-omics dataset that combines measurements from up to 10 different omics technologies applied to the same biological system, namely the well-studied mouse pre-B-cell differentiation. STATegra includes high-throughput measurements of chromatin structure, gene expression, proteomics and metabolomics, and it is complemented with single-cell data. To our knowledge, the STATegra collection is the most diverse multi-omics dataset describing a dynamic biological system.

Dramatic Remodeling of the Gut Microbiome Around Parturition and Its Relationship With Host Serum Metabolic Changes in Sows

Perinatal care is important in mammals due to its contribution to fetal growth, maternal health, and lactation. Substantial changes in host hormones, metabolism, and immunity around the parturition period may be accompanied by alterations in the gut microbiome. However, to our knowledge, changes in the gut microbiome and their contribution to the shifts in host metabolism around parturition have not been investigated in pigs. Furthermore, pigs are an ideal biomedical model for studying the interactions of the gut microbiota with host metabolism, due to the ease of controlling feeding conditions. Here we report dramatic remodeling of the gut microbiota and the potential functional capacity during the late stages of pregnancy (5 days before parturition, LP) to postpartum (within 6 h after delivery, PO) in both experimental and validated populations of sows (n = 107). The richness of bacteria in the gut of both pregnant and delivery sows significantly decreased, whilst the β-diversity dramatically expanded. The ratio of Bacteroidetes to Firmicutes, and the relative abundance of Prevotella significantly decreased, whilst the relative abundance of the predominant genus Lactobacillus significantly increased from LP to PO state. The predicted functional capacities of the gut microbiome related to amino acid metabolism, the metabolism of cofactors and vitamins, and glycan biosynthesis were significantly decreased from LP to PO state. However, the abundance of the functional capacities associated with carbohydrate and lipid metabolism were increased. Consistent with these changes, serum metabolites enriched at the LP stage were associated with the metabolism of amino acids and vitamins. In contrast, metabolites enriched at the PO stage were related to lipid metabolism. We further identified that the richness and β-diversity of the gut microbiota and the abundance of Lactobacillus accounted for shifts in the levels of bile acid metabolites associated with lipid metabolism. The results suggest that host-microbiota interactions during the perinatal period impact host metabolism. These benefit the lactation of sows by providing energy from lipid metabolism for milk production.

Cooperation of cancer drivers with regulatory germline variants shapes clinical outcomes

Pediatric malignancies including Ewing sarcoma (EwS) feature a paucity of somatic alterations except for pathognomonic driver-mutations that cannot explain overt variations in clinical outcome. Here, we demonstrate in EwS how cooperation of dominant oncogenes and regulatory germline variants determine tumor growth, patient survival and drug response. Binding of the oncogenic EWSR1-FLI1 fusion transcription factor to a polymorphic enhancer-like DNA element controls expression of the transcription factor MYBL2 mediating these phenotypes. Whole-genome and RNA sequencing reveals that variability at this locus is inherited via the germline and is associated with variable inter-tumoral MYBL2 expression. High MYBL2 levels sensitize EwS cells for inhibition of its upstream activating kinase CDK2 in vitro and in vivo, suggesting MYBL2 as a putative biomarker for anti-CDK2-therapy. Collectively, we establish cooperation of somatic mutations and regulatory germline variants as a major determinant of tumor progression and highlight the importance of integrating the regulatory genome in precision medicine.

Gene expression and immunohistochemical analyses identify SOX2 as major risk factor for overall survival and relapse in Ewing sarcoma patients

BACKGROUND

Up to 30-40% of Ewing sarcoma (EwS) patients with non-metastatic disease develop local or metastatic relapse within a time span of 2-10 years. This is in part caused by the absence of prognostic biomarkers that can identify high-risk patients and thus assign them to risk-adapted monitoring and treatment regimens. Since cancer stemness has been associated with tumour relapse and poor patient outcomes, we investigated in the current study the prognostic potential SOX2 (sex determining region Y box 2) - a major transcription factor involved in development and stemness - which was previously described to contribute to the undifferentiated phenotype of EwS.

METHODS

Two independent patient cohorts, one consisting of 189 retrospectively collected EwS tumours with corresponding mRNA expression data (test-cohort) and the other consisting of 141 prospectively collected formalin-fixed and paraffin-embedded resected tumours (validation and cohort), were employed to analyse SOX2 expression levels through DNA microarrays or immunohistochemistry, respectively, and to compare them with clinical parameters and patient outcomes. Two methods were employed to test the validity of the results at both the mRNA and protein levels.

FINDINGS

Both cohorts showed that only a subset of EwS patients (16-20%) expressed high SOX2 mRNA or protein levels, which significantly correlated with poor overall survival. Multivariate analyses of our validation-cohort revealed that high SOX2 expression represents a major risk-factor for poor survival (HR = 3·19; 95%CI 1·74-5·84; p < 0·01) that is independent from metastasis and other known clinical risk-factors at the time of diagnosis. Univariate analyses demonstrated that SOX2-high expression was correlated with tumour relapse (p = 0·002). The median first relapse was at 14·7 months (range: 3·5-180·7).

INTERPRETATION

High SOX2 expression constitutes an independent prognostic biomarker for EwS patients with poor outcomes. This may help to identify patients with localised disease who are at high risk for tumour relapse within the first two years after diagnosis.

FUNDING

The laboratory of T. G. P. Grünewald is supported by grants from the 'Verein zur Förderung von Wissenschaft und Forschung an der Medizinischen Fakultät der LMU München (WiFoMed)', by LMU Munich's Institutional Strategy LMUexcellent within the framework of the German Excellence Initiative, the 'Mehr LEBEN für krebskranke Kinder - Bettina-Bräu-Stiftung', the Walter Schulz Foundation, the Wilhelm Sander-Foundation (2016.167.1), the Friedrich-Baur foundation, the Matthias-Lackas foundation, the Barbara & Hubertus Trettner foundation, the Dr. Leopold & Carmen Ellinger foundation, the Gert & Susanna Mayer foundation, the Deutsche Forschungsgemeinschaft (DFG 391665916), and by the German Cancer Aid (DKH-111886 and DKH-70112257). J. Li was supported by a scholarship of the China Scholarship Council (CSC), J. Musa was supported by a scholarship of the Kind-Philipp foundation, and T. L. B. Hölting by a scholarship of the German Cancer Aid. M. F. Orth and M. M. L. Knott were supported by scholarships of the German National Academic Foundation. G. Sannino was supported by a scholarship from the Fritz-Thyssen Foundation (FTF-40.15.0.030MN). The work of U. Dirksen is supported by grants from the German Cancer Aid (DKH-108128, DKH-70112018, and DKH-70113419), the ERA-Net-TRANSCAN consortium (project number 01KT1310), and Euro Ewing Consortium (EEC, project number EU-FP7 602,856), both funded under the European Commission Seventh Framework Program FP7-HEALTH (http://cordis.europa.eu/), the Barbara & Hubertus Trettner foundation, and the Gert & Susanna Mayer foundation. G. Hardiman was supported by grants from the National Science Foundation (SC EPSCoR) and National Institutes of Health (U01-DA045300). The laboratory of J. Alonso was supported by Instituto de Salud Carlos III (PI12/00816; PI16CIII/00026); Asociación Pablo Ugarte (TPY-M 1149/13; TRPV 205/18), ASION (TVP 141/17), Fundación Sonrisa de Alex & Todos somos Iván (TVP 1324/15).

TRNA-derived fragments as sex-dependent circulating candidate biomarkers for Parkinson's disease

INTRODUCTION

Parkinson's Disease (PD) is diagnosed clinically. Reliable non-invasive PD biomarkers are actively sought. Transfer RNAs produce short non-coding RNAs, the tRNA-derived fragments (tRF). tRF have been shown to play diverse roles, including in amyotrophic lateral sclerosis, and the response to ischemic stroke. Rich tRF populations are being reported in biofluids. We explored the possibility that tRF can serve as non-invasive biomarkers for PD.

METHODS

We collected existing RNA-seq samples and re-analyzed a total of 254 legacy datasets from 3 previous studies, from male and female PD patients and controls that belong to three categories: prefrontal cortex samples from 29 patients and 33 controls; cerebrospinal fluid (CSF) samples from 63 patients and 64 controls; and, serum samples from 34 patients and 31 controls. First, we identified tRF exhaustively and deterministically in every dataset. Second, we determined tRF that are differentially abundant (DA) between PD and control samples, using uncorrected t-tests. Lastly, we assessed all the DA tRF from the previous step with Partial Least Squares - Discriminant Analysis (PLS-DA) to stringently sub-select tRF that can distinguish PD patients from controls.

RESULTS

We show that PLS-DA identified tRF from prefrontal cortex, CSF, and serum that can distinguish PD patients from controls. A handful of identified tRF were previously investigated in neurological contexts. Signatures built from relatively few tRF suffice to distinguish PD from control in each category of samples with high sensitivity (89-100%) and specificity (79-98%).

CONCLUSION

tRF-based signatures are promising candidates that warrant further evaluation as non-invasive PD biomarkers.

Survival marker genes of colorectal cancer derived from consistent transcriptomic profiling

Background

Identification of biomarkers associated with the prognosis of different cancer subtypes is critical to achieve better therapeutic assistance. In colorectal cancer (CRC) the discovery of stable and consistent survival markers remains a challenge due to the high heterogeneity of this class of tumors. In this work, we identified a new set of gene markers for CRC associated to prognosis and risk using a large unified cohort of patients with transcriptomic profiles and survival information.

Results

We built an integrated dataset with 1273 human colorectal samples, which provides a homogeneous robust framework to analyse genome-wide expression and survival data. Using this dataset we identified two sets of genes that are candidate prognostic markers for CRC in stages III and IV, showing either up-regulation correlated with poor prognosis or up-regulation correlated with good prognosis. The top 10 up-regulated genes found as survival markers of poor prognosis (i.e. low survival) were: DCBLD2, PTPN14, LAMP5, TM4SF1, NPR3, LEMD1, LCA5, CSGALNACT2, SLC2A3 and GADD45B. The stability and robustness of the gene survival markers was assessed by cross-validation, and the best-ranked genes were also validated with two external independent cohorts: one of microarrays with 482 samples; another of RNA-seq with 269 samples. Up-regulation of the top genes was also proved in a comparison with normal colorectal tissue samples. Finally, the set of top 100 genes that showed overexpression correlated with low survival was used to build a CRC risk predictor applying a multivariate Cox proportional hazards regression analysis. This risk predictor yielded an optimal separation of the individual patients of the cohort according to their survival, with a p-value of 8.25e-14 and Hazard Ratio 2.14 (95% CI: 1.75–2.61).

Conclusions

The results presented in this work provide a solid rationale for the prognostic utility of a new set of genes in CRC, demonstrating their potential to predict colorectal tumor progression and evolution towards poor survival stages. Our study does not provide a fixed gene signature for prognosis and risk prediction, but instead proposes a robust set of genes ranked according to their predictive power that can be selected for additional tests with other CRC clinical cohorts.

Electronic supplementary material

The online version of this article (10.1186/s12864-018-5193-9) contains supplementary material, which is available to authorized users.

MetaboClust: Using interactive time-series cluster analysis to relate metabolomic data with perturbed pathways

MOTIVATION

Modern analytical techniques such as LC-MS, GC-MS and NMR are increasingly being used to study the underlying dynamics of biological systems by tracking changes in metabolite levels over time. Such techniques are capable of providing information on large numbers of metabolites simultaneously, a feature that is exploited in non-targeted studies. However, since the dynamics of specific metabolites are unlikely to be known a priori this presents an initial subjective challenge as to where the focus of the investigation should be. Whilst a number of feed-forward software tools are available for manipulation of metabolomic data, no tool centralizes on clustering and focus is typically directed by a workflow that is chosen in advance.

RESULTS

We present an interactive approach to time-course analyses and a complementary implementation in a software package, MetaboClust. This is presented through the analysis of two LC-MS time-course case studies on plants (Medicago truncatula and Alopecurus myosuroides). We demonstrate a dynamic, user-centric workflow to clustering with intrinsic visual feedback at all stages of analysis. The software is used to apply data correction, generate the time-profiles, perform exploratory statistical analysis and assign tentative metabolite identifications. Clustering is used to group metabolites in an unbiased manner, allowing pathway analysis to score metabolic pathways, based on their overlap with clusters showing interesting trends.

Consensus molecular subtype classification of colorectal adenomas

Consensus molecular subtyping is an RNA expression‐based classification system for colorectal cancer (CRC). Genomic alterations accumulate during CRC pathogenesis, including the premalignant adenoma stage, leading to changes in RNA expression. Only a minority of adenomas progress to malignancies, a transition that is associated with specific DNA copy number aberrations or microsatellite instability (MSI). We aimed to investigate whether colorectal adenomas can already be stratified into consensus molecular subtype (CMS) classes, and whether specific CMS classes are related to the presence of specific DNA copy number aberrations associated with progression to malignancy. RNA sequencing was performed on 62 adenomas and 59 CRCs. MSI status was determined with polymerase chain reaction‐based methodology. DNA copy number was assessed by low‐coverage DNA sequencing (n = 30) or array‐comparative genomic hybridisation (n = 32). Adenomas were classified into CMS classes together with CRCs from the study cohort and from The Cancer Genome Atlas (n = 556), by use of the established CMS classifier. As a result, 54 of 62 (87%) adenomas were classified according to the CMS. The CMS3 ‘metabolic subtype’, which was least common among CRCs, was most prevalent among adenomas (n = 45; 73%). One of the two adenomas showing MSI was classified as CMS1 (2%), the ‘MSI immune’ subtype. Eight adenomas (13%) were classified as the ‘canonical’ CMS2. No adenomas were classified as the ‘mesenchymal’ CMS4, consistent with the fact that adenomas lack invasion‐associated stroma. The distribution of the CMS classes among adenomas was confirmed in an independent series. CMS3 was enriched with adenomas at low risk of progressing to CRC, whereas relatively more high‐risk adenomas were observed in CMS2. We conclude that adenomas can be stratified into the CMS classes. Considering that CMS1 and CMS2 expression signatures may mark adenomas at increased risk of progression, the distribution of the CMS classes among adenomas is consistent with the proportion of adenomas expected to progress to CRC. © 2018 The Authors. The Journal of Pathology published by John Wiley & Sons Ltd on behalf of Pathological Society of Great Britain and Ireland.