Generation Of Biological Hypotheses Research Articles

Machine learning for hypothesis generation in biology and medicine: exploring the latent space of neuroscience and developmental bioelectricity

FieldSHIFT uses in-context learning to translate neuroscience abstracts into developmental biology abstracts based on example concept mappings, creating new research hypotheses at scale.

Adding a Chemical Biology Twist to CRISPR Screening.

In less than a decade, CRISPR screening has revolutionized forward genetics and cell and molecular biology. Advances in screening technologies, including sgRNA libraries, Cas9-expressing cell lines, and streamlined sequencing pipelines, have democratized pooled CRISPR screens at genome-wide scale. Initially, many such screens were survival-based, identifying essential genes in physiological or perturbed processes. With the application of new chemical biology tools to CRISPR screening, the phenotypic space is no longer limited to live/dead selection or screening for levels of conventional fluorescent protein reporters. Further, the resolution has been increased from cell populations to single cells or even the subcellular level. We highlight advances in pooled CRISPR screening, powered by chemical biology, that have expanded phenotypic space, resolution, scope, and scalability as well as strengthened the CRISPR/Cas enzyme toolkit to enable biological hypothesis generation and discovery.

MoSBi: Automated signature mining for molecular stratification and subtyping

The improving access to increasing amounts of biomedical data provides completely new chances for advanced patient stratification and disease subtyping strategies. This requires computational tools that produce uniformly robust results across highly heterogeneous molecular data. Unsupervised machine learning methodologies are able to discover de novo patterns in such data. Biclustering is especially suited by simultaneously identifying sample groups and corresponding feature sets across heterogeneous omics data. The performance of available biclustering algorithms heavily depends on individual parameterization and varies with their application. Here, we developed MoSBi (molecular signature identification using biclustering), an automated multialgorithm ensemble approach that integrates results utilizing an error model-supported similarity network. We systematically evaluated the performance of 11 available and established biclustering algorithms together with MoSBi. For this, we used transcriptomics, proteomics, and metabolomics data, as well as synthetic datasets covering various data properties. Profiting from multialgorithm integration, MoSBi identified robust group and disease-specific signatures across all scenarios, overcoming single algorithm specificities. Furthermore, we developed a scalable network-based visualization of bicluster communities that supports biological hypothesis generation. MoSBi is available as an R package and web service to make automated biclustering analysis accessible for application in molecular sample stratification.

Metabolomics and Microbiomics: New Potential Strategies in Chronic Pain Syndrome.

Chronic pain affects almost 20% of the European adult population and it significantly reduces patients’ quality of life. Chronic pain is considered a multidimensional experience determined by the interaction of several genetic and environmental factors. The effect of specific genetic contributions is often unclear, and the interpretation of the results from studies focused on genetic influences on pain has been complicated by the existence of multiple pain phenotypes. A step forward from genetics could be given by the application of metabolomics and microbiomics tools. Metabolomics is a powerful approach for hypothesis generation in biology, and it aims to analyze low molecular weight compounds, either metabolic intermediates or metabolic end-products, resulting from human or microbial metabolism. Microbiomics is a fast-growing field in which all the microbes are examined together, and as a result, its perturbation may indicate the development of chronic diseases. By applying these methodologies for the study of chronic pain, several differences have been identified. The alteration of the choline-PAF pathway is an intriguing finding recognized by several groups. In our opinion, metabolomics and microbiomics techniques will allow significant progress into the medical field. Patients may benefit from the possibility of being stratified and classified based on their metabolic and microbial profile, which, in the next future, may lead to personalized therapy.

Computational Interspecies Translation Between Alzheimer's Disease Mouse Models and Human Subjects Identifies Innate Immune Complement, TYROBP, and TAM Receptor Agonist Signatures, Distinct From Influences of Aging.

Mouse models are vital for preclinical research on Alzheimer’s disease (AD) pathobiology. Many traditional models are driven by autosomal dominant mutations identified from early onset AD genetics whereas late onset and sporadic forms of the disease are predominant among human patients. Alongside ongoing experimental efforts to improve fidelity of mouse model representation of late onset AD, a computational framework termed Translatable Components Regression (TransComp-R) offers a complementary approach to leverage human and mouse datasets concurrently to enhance translation capabilities. We employ TransComp-R to integratively analyze transcriptomic data from human postmortem and traditional amyloid mouse model hippocampi to identify pathway-level signatures present in human patient samples yet predictive of mouse model disease status. This method allows concomitant evaluation of datasets across different species beyond observational seeking of direct commonalities between the species. Additional linear modeling focuses on decoupling disease signatures from effects of aging. Our results elucidated mouse-to-human translatable signatures associated with disease: excitatory synapses, inflammatory cytokine signaling, and complement cascade- and TYROBP-based innate immune activity; these signatures all find validation in previous literature. Additionally, we identified agonists of the Tyro3 / Axl / MerTK (TAM) receptor family as significant contributors to the cross-species innate immune signature; the mechanistic roles of the TAM receptor family in AD merit further dedicated study. We have demonstrated that TransComp-R can enhance translational understanding of relationships between AD mouse model data and human data, thus aiding generation of biological hypotheses concerning AD progression and holding promise for improved preclinical evaluation of therapies.

Generation of biological hypotheses by functional imaging links tumor hypoxia to radiation induced tissue inflammation/glucose uptake in head and neck cancer

Background and purposePositron emission tomography (PET) is a functional imaging modality which is able to deliver tracer specific biological information, e.g. about glucose uptake, inflammation or hypoxia of tumors. We performed a proof-of-principle study that used different tracers and expanded the analytical scope to non-tumor structures to evaluate tumor-host interactions. Materials and methodsBased on a previously reported prospective imaging study on 50 patients treated with curative intent chemoradiation (CRT) for head and neck squamous cell carcinoma, PET-based hypoxia and normal tissue inflammation measured by repeat 18F-fluoromisonidazole (FMISO) PET and 18F-fluorodesoxyglucose (FDG) PET, respectively, were correlated using the Spearman correlation coefficient R. PET parameters determined before and during CRT (week 1, 2 and 5), were associated with local tumor control and overall survival. ResultsTumor hypoxia at all measured times showed an inverse correlation with mid-treatment FDG-uptake of non-tumor affected oral (sub-)mucosa with R values between −0.35 and −0.6 (all p < 0.05). Mucosal FDG-uptake and mucosal hypoxia correlated positively but weaker (R values between 0.2 and 0.45). More tumor hypoxia in FMISO-PET (week 2) and less FDG-uptake of (sub-)mucosa in FDG-PET (week 4) were significantly associated with worse LC (FMISO TBRpeak: HR = 1.72, p = 0.030; FDG SUVmean: HR = 0.23, p = 0.025) and OS (FMISO TBRpeak: HR = 1.71, p = 0.007; FDG SUVmean: HR = 0.30, p = 0.003). Multivariable models including both parameters showed improved performance, suggesting that these modalities still bear distinct biological information despite their strong inter-correlation. ConclusionWe report first clinical evidence that tumor hypoxia is inversely correlated with increased FDG-uptake during radiation, potentially expressing inflammation. This observation merits further research and may have important implication for future research on tumor hypoxia and radio-immunology. Our study demonstrates that functional imaging can be utilized to assess complex tumor-host interactions and generate novel biological insights in vivo vero.

Functional network analysis reveals an immune tolerance mechanism in cancer

We present a technique to construct a simplification of a feature network which can be used for interactive data exploration, biological hypothesis generation, and the detection of communities or modules of cofunctional features. These are modules of features that are not necessarily correlated, but nevertheless exhibit common function in their network context as measured by similarity of relationships with neighboring features. In the case of genetic networks, traditional pathway analyses tend to assume that, ideally, all genes in a module exhibit very similar function, independent of relationships with other genes. The proposed technique explicitly relaxes this assumption by employing the comparison of relational profiles. For example, two genes which always activate a third gene are grouped together even if they never do so concurrently. They have common, but not identical, function. The comparison is driven by an average of a certain computationally efficient comparison metric between Gaussian mixture models. The method has its basis in the local connection structure of the network and the collection of joint distributions of the data associated with nodal neighborhoods. It is benchmarked on networks with known community structures. As the main application, we analyzed the gene regulatory network in lung adenocarcinoma, finding a cofunctional module of genes including the pregnancy-specific glycoproteins (PSGs). About 20% of patients with lung, breast, uterus, and colon cancer in The Cancer Genome Atlas (TCGA) have an elevated PSG+ signature, with associated poor group prognosis. In conjunction with previous results relating PSGs to tolerance in the immune system, these findings implicate the PSGs in a potential immune tolerance mechanism of cancers.

GREG-studying transcriptional regulation using integrative graph databases.

A gene regulatory process is the result of the concerted action of transcription factors, co-factors, regulatory non-coding RNAs (ncRNAs) and chromatin interactions. Therefore, the combination of protein–DNA, protein–protein, ncRNA–DNA, ncRNA–protein and DNA–DNA data in a single graph database offers new possibilities regarding generation of biological hypotheses. GREG (The Gene Regulation Graph Database) is an integrative database and web resource that allows the user to visualize and explore the network of all above-mentioned interactions for a query transcription factor, long non-coding RNA, genomic range or DNA annotation, as well as extracting node and interaction information, identifying connected nodes and performing advanced graphical queries directly on the regulatory network, in a simple and efficient way. In this article, we introduce GREG together with some application examples (including exploratory research of Nanog’s regulatory landscape and the etiology of chronic obstructive pulmonary disease), which we use as a demonstration of the advantages of using graph databases in biomedical research. Database URL: https://mora-lab.github.io/projects/greg.html, www.moralab.science/GREG/

Extremely-randomized-tree-based Prediction of N6-Methyladenosine Sites in Saccharomyces cerevisiae.

Introduction N6-methyladenosine (m6A) is one of the most common post-transcriptional modifications in RNA, which has been related to several biological processes. The accurate prediction of m6A sites from RNA sequences is one of the challenging tasks in computational biology. Several computational methods utilizing machine-learning algorithms have been proposed that accelerate in silico screening of m6A sites, thereby drastically reducing the experimental time and labor costs involved.Methodology In this study, we proposed a novel computational predictor termed ERT-m6Apred, for the accurate prediction of m6A sites. To identify the feature encodings with more discriminative capability, we applied a two-step feature selection technique on seven different feature encodings and identified the corresponding optimal feature set.Results Subsequently, performance comparison of the corresponding optimal feature set-based extremely randomized tree model revealed that Pseudo k-tuple composition encoding, which includes 14 physicochemical properties significantly outperformed other encodings. Moreover, ERT-m6Apred achieved an accuracy of 78.84% during cross-validation analysis, which is comparatively better than recently reported predictors.Conclusion In summary, ERT-m6Apred predicts Saccharomyces cerevisiae m6A sites with higher accuracy, thus facilitating biological hypothesis generation and experimental validations.

SIMLIN: a bioinformatics tool for prediction of S-sulphenylation in the\xa0human proteome based on multi-stage ensemble-learning models

BackgroundS-sulphenylation is a ubiquitous protein post-translational modification (PTM) where an S-hydroxyl (−SOH) bond is formed via the reversible oxidation on the Sulfhydryl group of cysteine (C). Recent experimental studies have revealed that S-sulphenylation plays critical roles in many biological functions, such as protein regulation and cell signaling. State-of-the-art bioinformatic advances have facilitated high-throughput in silico screening of protein S-sulphenylation sites, thereby significantly reducing the time and labour costs traditionally required for the experimental investigation of S-sulphenylation.ResultsIn this study, we have proposed a novel hybrid computational framework, termed SIMLIN, for accurate prediction of protein S-sulphenylation sites using a multi-stage neural-network based ensemble-learning model integrating both protein sequence derived and protein structural features. Benchmarking experiments against the current state-of-the-art predictors for S-sulphenylation demonstrated that SIMLIN delivered competitive prediction performance. The empirical studies on the independent testing dataset demonstrated that SIMLIN achieved 88.0% prediction accuracy and an AUC score of 0.82, which outperforms currently existing methods.ConclusionsIn summary, SIMLIN predicts human S-sulphenylation sites with high accuracy thereby facilitating biological hypothesis generation and experimental validation. The web server, datasets, and online instructions are freely available at http://simlin.erc.monash.edu/ for academic purposes.

PlantSimLab - a modeling and simulation web tool for plant biologists

BackgroundAt the molecular level, nonlinear networks of heterogeneous molecules control many biological processes, so that systems biology provides a valuable approach in this field, building on the integration of experimental biology with mathematical modeling. One of the biggest challenges to making this integration a reality is that many life scientists do not possess the mathematical expertise needed to build and manipulate mathematical models well enough to use them as tools for hypothesis generation. Available modeling software packages often assume some modeling expertise. There is a need for software tools that are easy to use and intuitive for experimentalists.ResultsThis paper introduces PlantSimLab, a web-based application developed to allow plant biologists to construct dynamic mathematical models of molecular networks, interrogate them in a manner similar to what is done in the laboratory, and use them as a tool for biological hypothesis generation. It is designed to be used by experimentalists, without direct assistance from mathematical modelers.ConclusionsMathematical modeling techniques are a useful tool for analyzing complex biological systems, and there is a need for accessible, efficient analysis tools within the biological community. PlantSimLab enables users to build, validate, and use intuitive qualitative dynamic computer models, with a graphical user interface that does not require mathematical modeling expertise. It makes analysis of complex models accessible to a larger community, as it is platform-independent and does not require extensive mathematical expertise.

Phosphorus-Based Probes as Molecular Tools for Proteome Studies: Recent Advances in Probe Development and Applications.

Studies on the human proteome have engaged diverse techniques; however, none of them represent a predominant approach. Chemical biology has made a major contribution to our understanding of human biology, stimulating the generation of biological hypotheses. Tools such as functional probes have advanced studies on biological mechanisms and helped in elucidating off-target reactivity and potential toxicities of drugs and drug candidates. Here, we accentuate the recent developments in the design and applications of phosph(on)ate-containing probes. Phosphate esters and anhydrides are present in a number of vital cell constituents, and their significance can be reflected by a number of biological processes that involve phosphorus-bearing molecules. We discuss the use of phosph(on)ate-derived probes for (1) the identification of phosphate-requiring enzymes, their substrates, interacting partners; (2) developing screening assays; and (3) their potential as diagnostics. Limitations that as yet need to be overcome and possible measures to be undertaken will also be addressed.

Biomarker Detection and Categorization in Ribonucleic Acid Sequencing Meta-Analysis Using Bayesian Hierarchical Models

Meta-analysis combining multiple transcriptomic studies increases statistical power and accuracy in detecting differentially expressed genes. As the next-generation sequencing experiments become mature and affordable, increasing number of RNA-seq datasets are available in the public domain. The count-data based technology provides better experimental accuracy, reproducibility and ability to detect low-expressed genes. A naive approach to combine multiple RNA-seq studies is to apply differential analysis tools such as edgeR and DESeq to each study and then combine the summary statistics of p-values or effect sizes by conventional meta-analysis methods. Such a two-stage approach loses statistical power, especially for genes with short length or low expression abundance. In this paper, we propose a full Bayesian hierarchical model (namely, BayesMetaSeq) for RNA-seq meta-analysis by modelling count data, integrating information across genes and across studies, and modelling potentially heterogeneous differential signals across studies via latent variables. A Dirichlet process mixture (DPM) prior is further applied on the latent variables to provide categorization of detected biomarkers according to their differential expression patterns across studies, facilitating improved interpretation and biological hypothesis generation. Simulations and a real application on multi-brain-region HIV-1 transgenic rats demonstrate improved sensitivity, accuracy and biological findings of the proposed method.

Hypothesis Generation in Biology

Helping students understand and generate appropriate hypotheses and test their subsequent predictions – in science in general and biology in particular – should be at the core of teaching the nature of science. However, there is much confusion among students and teachers about the difference between hypotheses and predictions. Here, I present evidence of the problem and describe steps that scientists actually follow when employing scientific reasoning strategies. This is followed by a proposed solution for helping students effectively explore this important aspect of the nature of science.

MCA: Multiresolution Correlation Analysis, a graphical tool for subpopulation identification in single-cell gene expression data.

BackgroundBiological data often originate from samples containing mixtures of subpopulations, corresponding e.g. to distinct cellular phenotypes. However, identification of distinct subpopulations may be difficult if biological measurements yield distributions that are not easily separable.ResultsWe present Multiresolution Correlation Analysis (MCA), a method for visually identifying subpopulations based on the local pairwise correlation between covariates, without needing to define an a priori interaction scale. We demonstrate that MCA facilitates the identification of differentially regulated subpopulations in simulated data from a small gene regulatory network, followed by application to previously published single-cell qPCR data from mouse embryonic stem cells. We show that MCA recovers previously identified subpopulations, provides additional insight into the underlying correlation structure, reveals potentially spurious compartmentalizations, and provides insight into novel subpopulations.ConclusionsMCA is a useful method for the identification of subpopulations in low-dimensional expression data, as emerging from qPCR or FACS measurements. With MCA it is possible to investigate the robustness of covariate correlations with respect subpopulations, graphically identify outliers, and identify factors contributing to differential regulation between pairs of covariates. MCA thus provides a framework for investigation of expression correlations for genes of interests and biological hypothesis generation.

From -omics to personalized medicine in nephrology: integration is the key

Large-scale gene, protein and metabolite measurements ('omics') have driven the resolution of biology to an unprecedented high definition. Passing from reductionism to a system-oriented perspective, medical research will take advantage of these high-throughput technologies unveiling their full potential. Integration is the key to decoding the underlying principles that govern the complex functions of living systems. Extensive computational support and statistical modelling is needed to manage and connect the -omic data sets but this, in turn, is speeding up the hypothesis generation in biology enormously and yielding a deep insight into the pathophysiology. This systems biology approach will transform diagnostic and therapeutic strategies with the discovery of novel biomarkers that will enable a predictive and preventive medicine leading to personalized medicine.

Tools for protein-protein interaction network analysis in cancer research

As cancer is a complex disease, the representation of a malignant cell as a protein-protein interaction network (PPIN) and its subsequent analysis can provide insight into the behaviour of cancer cells and lead to the discovery of new biomarkers. The aim of this review is to help life-science researchers without previous computer programming skills to extract meaningful biological information from such networks, taking advantage of easy-to-use, public bioinformatics tools. It is structured in four parts: the first section describes the pipeline of consecutive steps from network construction to biological hypothesis generation. The second part provides a repository of public, user-friendly tools for network construction, visualisation and analysis. Two different and complementary approaches of network analysis are presented: the topological approach studies the network as a whole by means of structural graph theory, whereas the global approach divides the PPIN into sub-graphs, or modules. In section three, some concepts and tools regarding heterogeneous molecular data integration through a PPIN are described. Finally, the fourth part is an example of how to extract meaningful biological information from a colorectal cancer PPIN using some of the described tools.

Science Knowledge and Scientific Emotion during the Generation of Biological Hypotheses in High School Students

Science Knowledge and Scientific Emotion during the Generation of Biological Hypotheses in High School Students

Improvement of the quantitative differential metabolome pipeline for gas chromatography-mass spectrometry data by automated reliable peak selection

Recent advances in metabolomics technology have enabled large-scale comprehensive analyses of metabolites, but the throughput of data processing of non-targeted, quantitative differential analyses is very low. It is crucial to solve this problem to generate biological hypotheses from a large-scale dataset. To improve the analysis of metabolite data, we focused on the processing of quantitative differential analysis after multiple peak alignment. We have developed a program named FAQuant that automatically selects reliable peaks from each chromatogram, quantifies the mean of peak intensity to compare between sample groups, and selects the peaks with differences in accumulation of metabolites. This program was incorporated into a quantitative differential metabolome pipeline as a module to improve the throughput of gas chromatography-mass spectrometry dataset analysis. As a result, the module incorporation largely reduced the total processing time. Furthermore, differential analysis of metabolites in soybean (Glycine max) cultivars was demonstrated by use of the system. This system might facilitate biological hypothesis generation from large-scale comparative metabolome analysis.

Changes in brain activation induced by the training of hypothesis generation skills: An fMRI study

The aim of the present study is to investigate the learning-related changes in brain activation induced by the training of hypothesis generation skills regarding biological phenomena. Eighteen undergraduate participants were scanned twice with functional magnetic resonance imaging (fMRI) before and after training over a period of 2 months. The experimental group underwent eight biological hypothesis generation training programs, but the control group was not given any during the 2-month period. The results showed that the left frontal gyri, the cingulate gyrus, and the cuneus were activated during hypothesis generation. In addition, the brain activation of the trained group increased in the left inferior and the superior frontal gyri, which are related to working memory load and higher-order inferential processes. However, the activation after training decreased in the occipito-parietal route, which is associated with the perception and the analysis processes of visual information. Furthermore, the results have suggested that the dorsolateral prefrontal cortex (DLPFC) region is the critical area in the training of hypothesis generation skills.