iPS-cell-derived microglia promote brain organoid maturation via cholesterol transfer.

Microglia are specialized brain-resident macrophages that arise from primitive macrophages colonizing the embryonic brain1. Microglia contribute to multiple aspects of brain development, but their precise roles in the early human brain remain poorly understood owing to limited access to relevant tissues2-6. The generation of brain organoids from human induced pluripotent stem cells recapitulates some key features of human embryonic brain development7-10. However, current approaches do not incorporate microglia or address their role in organoid maturation11-21. Here we generated microglia-sufficient brain organoids by coculturing brain organoids with primitive-like macrophages generated from the same human induced pluripotent stem cells (iMac)22. In organoid cocultures, iMac differentiated into cells with microglia-like phenotypes and functions (iMicro) and modulated neuronal progenitor cell (NPC) differentiation, limiting NPC proliferation and promoting axonogenesis. Mechanistically, iMicro contained high levels of PLIN2+ lipid droplets that exported cholesterol and its esters, which were taken up by NPCs in the organoids. We also detected PLIN2+ lipid droplet-loaded microglia in mouse and human embryonic brains. Overall, our approach substantially advances current human brain organoid approaches by incorporating microglial cells, as illustrated by the discovery of a key pathway of lipid-mediated crosstalk between microglia and NPCs that leads to improved neurogenesis.

Reciprocal signalling by Notch-Collagen V-CALCR retains muscle stem cells in their niche.

The cell microenvironment, which is critical for stem cell maintenance, contains both cellular and non-cellular components, including secreted growth factors and the extracellular matrix1-3. Although Notch and other signalling pathways have previously been reported to regulate quiescence of stem cells4-9, the composition and source of molecules that maintain the stem cell niche remain largely unknown. Here we show that adult muscle satellite (stem) cells in mice produce extracellular matrix collagens to maintain quiescence in a cell-autonomous manner. Using chromatin immunoprecipitation followed by sequencing, we identified NOTCH1/RBPJ-bound regulatory elements adjacent to specific collagen genes, the expression of which is deregulated in Notch-mutant mice. Moreover, we show that Collagen V (COLV) produced by satellite cells is a critical component of the quiescent niche, as depletion of COLV by conditional deletion of the Col5a1 gene leads to anomalous cell cycle entry and gradual diminution of the stem cell pool. Notably, the interaction of COLV with satellite cells is mediated by the Calcitonin receptor, for which COLV acts as a surrogate local ligand. Systemic administration of a calcitonin derivative is sufficient to rescue the quiescence and self-renewal defects found in COLV-null satellite cells. This study reveals a Notch-COLV-Calcitonin receptor signalling cascade that maintains satellite cells in a quiescent state in a cell-autonomous fashion, and raises the possibility that similar reciprocal mechanisms act in diverse stem cell populations.

A new subtype of diffuse midline glioma, H3 K27 and BRAF/FGFR1 co-altered: a clinico-radiological and histomolecular characterisation.

Diffuse midline gliomas (DMG) H3 K27-altered are incurable grade 4 gliomas and represent a major challenge in neuro-oncology. This tumour type is now classified in four subtypes by the 2021 edition of the WHO Classification of the Central Nervous System (CNS) tumours. However, the H3.3-K27M subgroup still appears clinically and molecularly heterogeneous. Recent publications reported that rare patients presenting a co-occurrence of H3.3K27M with BRAF or FGFR1 alterations tended to have a better prognosis. To better study the role of these co-driver alterations, we assembled a large paediatric and adult cohort of 29 tumours H3K27-altered with co-occurring activating mutation in BRAF or FGFR1 as well as 31 previous cases from the literature. We performed a comprehensive histological, radiological, genomic, transcriptomic and DNA methylation analysis. Interestingly, unsupervised t-distributed Stochastic Neighbour Embedding (tSNE) analysis of DNA methylation profiles regrouped BRAFV600E and all but one FGFR1MUT DMG in a unique methylation cluster, distinct from the other DMG subgroups and also from ganglioglioma (GG) or high-grade astrocytoma with piloid features (HGAP). This new DMG subtype harbours atypical radiological and histopathological profiles with calcification and/or a solid tumour component both for BRAFV600E and FGFR1MUT cases. The analyses of a H3.3-K27M BRAFV600E tumour at diagnosis and corresponding in vitro cellular model showed that mutation in H3-3A was the first event in the oncogenesis. Contrary to other DMG, these tumours occur more frequently in the thalamus (70% for BRAFV600E and 58% for FGFR1MUT) and patients have a longer overall survival with a median above three years. In conclusion, DMG, H3 K27 and BRAF/FGFR1 co-altered represent a new subtype of DMG with distinct genotype/phenotype characteristics, which deserve further attention with respect to trial interpretation and patient management.

The genomic landscape of dysembryoplastic neuroepithelial tumours and a comprehensive analysis of recurrent cases.

AIMS: Dysembryoplastic neuroepithelial tumour (DNT) is a glioneuronal tumour that is challenging to diagnose, with a wide spectrum of histological features. Three histopathological patterns have been described: specific DNTs (both the simple form and the complex form) comprising the specific glioneuronal element, and also the non-specific/diffuse form which lacks it, and has unclear phenotype-genotype correlations with numerous differential diagnoses. METHODS: We used targeted methods (immunohistochemistry, fluorescence in situ hybridisation and targeted sequencing) and large-scale genomic methodologies including DNA methylation profiling to perform an integrative analysis to better characterise a large retrospective cohort of 82 DNTs, enriched for tumours that showed progression on imaging. RESULTS: We confirmed that specific DNTs are characterised by a single driver event with a high frequency of FGFR1 variants. However, a subset of DNA methylation-confirmed DNTs harbour alternative genomic alterations to FGFR1 duplication/mutation. We also demonstrated that a subset of DNTs sharing the same FGFR1 alterations can show in situ progression. In contrast to the specific forms, "non-specific/diffuse DNTs" corresponded to a heterogeneous molecular group encompassing diverse, newly-described, molecularly distinct entities. CONCLUSIONS: Specific DNT is a homogeneous group of tumours sharing characteristics of paediatric low-grade gliomas: a quiet genome with a recurrent genomic alteration in the RAS-MAPK signalling pathway, a distinct DNA methylation profile and a good prognosis but showing progression in some cases. The "non-specific/diffuse DNTs" subgroup encompasses various recently described histomolecular entities, such as PLNTY and diffuse astrocytoma, MYB or MYBL1 altered.

Radiogenomics of diffuse intrinsic pontine gliomas (DIPGs): correlation of histological and biological characteristics with multimodal MRI features.

OBJECTIVES: The diffuse intrinsic pontine gliomas (DIPGs) are now defined by the type of histone H3 mutated at lysine 27. We aimed to correlate the multimodal MRI features of DIPGs, H3K27M mutant, with their histological and molecular characteristics. METHODS: Twenty-seven treatment-naïve children with histopathologically confirmed DIPG H3K27M mutant were prospectively included. MRI performed prior to biopsy included multi-b-value diffusion-weighted imaging, ASL, and dynamic susceptibility contrast (DSC) perfusion imaging. The ADC and cerebral blood flow (CBF) and blood volume (CBV) were measured at the biopsy site. We assessed quantitative histological data, including microvascular density, nuclear density, and H3K27M-positive nuclear density. Gene expression profiling was also assessed in the samples. We compared imaging and histopathological data according to histone subgroup. We correlated MRI quantitative data with histological data and gene expression. RESULTS: H3.1K27M mutated tumors showed higher ADC values (median 3151 µm2/s vs 1741 µm2/s, p = 0.003), and lower perfusion values (DSC-rCBF median 0.71 vs 1.43, p = 0.002, and DSC-rCBV median 1.00 vs 1.71, p = 0.02) than H3.3K27M ones. They had similar microvascular and nuclear density, but lower H3K27M-positive nuclear density (p = 0.007). The DSC-rCBV was positively correlated to the H3K27M-positive nuclear density (rho = 0.74, p = 0.02). ADC values were not correlated with nuclear density nor perfusion values with microvascular density. The expression of gated channel activity-related genes tended to be inversely correlated with ADC values and positively correlated with DSC perfusion. CONCLUSIONS: H3.1K27M mutated tumors have higher ADC and lower perfusion values than H3.3K27M ones, without direct correlation with microvascular or nuclear density. This may be due to tissular edema possibly related to gated channel activity-related gene expression. KEY POINTS: • H3.1K27M mutant DIPG had higher apparent diffusion coefficient (p = 0.003), lower α (p = 0.048), and lower relative cerebral blood volume (p = 0.02) than H3.3K27M mutant DIPG at their biopsy sites. • Biopsy samples obtained within the tumor's enhancing portion showed higher microvascular density (p = 0.03) than samples obtained outside the tumor's enhancing portion, but similar H3K27M-positive nuclear density (p = 0.84). • Relative cerebral blood volume measured at the biopsy site was significantly correlated with H3K27M-positive nuclear density (rho = 0.74, p = 0.02).

Dynamic binding of RBPJ is determined by Notch signaling status.

Notch signaling plays crucial roles in mediating cell fate choices in all metazoans largely by specifying the transcriptional output of one cell in response to a neighboring cell. The DNA-binding protein RBPJ is the principle effector of this pathway in mammals and, together with the transcription factor moiety of Notch (NICD), regulates the expression of target genes. The prevalent view presumes that RBPJ statically occupies consensus binding sites while exchanging repressors for activators in response to NICD. We present the first specific RBPJ chromatin immunoprecipitation and high-throughput sequencing study in mammalian cells. To dissect the mode of transcriptional regulation by RBPJ and identify its direct targets, whole-genome binding profiles were generated for RBPJ; its coactivator, p300; NICD; and the histone H3 modifications H3 Lys 4 trimethylation (H3K4me3), H3 Lys 4 monomethylation (H3K4me1), and histone H3 Lys 27 acetylation (H3K27ac) in myogenic cells under active or inhibitory Notch signaling conditions. Our results demonstrate dynamic binding of RBPJ in response to Notch activation at essentially all sites co-occupied by NICD. Additionally, we identify a distinct set of sites where RBPJ recruits neither NICD nor p300 and binds DNA statically, irrespective of Notch activity. These findings significantly modify our views on how RBPJ and Notch signaling mediate their activities and consequently impact on cell fate decisions.

The histomolecular criteria established for adult anaplastic pilocytic astrocytoma are not applicable to the pediatric population.

Pilocytic astrocytoma (PA) is the most common pediatric glioma, arising from a single driver MAPK pathway alteration. Classified as a grade I tumor according to the 2016 WHO classification, prognosis is excellent with a 10-year survival rate > 95% after surgery. However, rare cases present with anaplastic features, including an unexpected high mitotic/proliferative index, thus posing a diagnostic and therapeutic challenge. Based on small histomolecular series and case reports, such tumors arising at the time of diagnosis or recurrence have been designated by many names including pilocytic astrocytoma with anaplastic features (PAAF). Recent DNA methylation-profiling studies performed mainly on adult cases have revealed that PAAF exhibit a specific methylation signature, thus constituting a distinct methylation class from typical PA [methylation class anaplastic astrocytoma with piloid features-(MC-AAP)]. However, the diagnostic and prognostic significance of MC-AAP remains to be determined in children. We performed an integrative work on the largest pediatric cohort of PAAF, defined according to strict criteria: morphology compatible with the diagnosis of PA, with or without necrosis, ≥ 4 mitoses for 2.3 mm2, and MAPK pathway alteration. We subjected 31 tumors to clinical, imaging, morphological and molecular analyses, including DNA methylation profiling. We identified only one tumor belonging to the MC-AAP (3%), the others exhibiting a methylation profile typical for PA (77%), IDH-wild-type glioblastoma (7%), and diffuse leptomeningeal glioneuronal tumor (3%), while three cases (10%) did not match to a known DNA methylation class. No significant outcome differences were observed between PAAF with necrosis versus no necrosis (p = 0.07), or with 4-6 mitoses versus 7 or more mitoses (p = 0.857). Our findings argue that the diagnostic histomolecular criteria established for anaplasia in adult PA are not of diagnostic or prognostic value in a pediatric setting. Further extensive and comprehensive integrative studies are necessary to accurately define this exceptional entity in children.

Correction to: International experience in the development of patient-derived xenograft models of diffuse intrinsic pontine glioma.

There are two errors and one omission in the original article. Author Gottardo's correct name is Nicholas G. Gottardo, author Hulleman's correct affiliation is no. 3 (VUMC, Amsterdam), and the Acknowledgements should include the following sentence: "We would like to thank Dr Angel Montero Carcaboso (Hospital Sant Joan de Deu, Barcelona, Spain) for generously supplying the HSJD-DIPG007 cells."

International experience in the development of patient-derived xenograft models of diffuse intrinsic pontine glioma.

PURPOSE: Diffuse intrinsic pontine glioma is the most aggressive form of high grade glioma in children with no effective therapies. There have been no improvements in survival in part due poor understanding of underlying biology, and lack of representative in vitro and in vivo models. Recently, it has been found feasible to use both biopsy and autopsy tumors to generate cultures and xenograft models. METHODS: To further model development, we evaluated the collective international experience from 8 collaborating centers to develop DIPG pre-clinical models from patient-derived autopsies and biopsies. Univariate and multivariate analysis was performed to determine key factors associated with the success of in vitro and in vivo PDX development. RESULTS: In vitro cultures were successfully established from 57% of samples (84.2% of biopsies and 38.2% of autopsies). Samples transferred in DMEM media were more likely to establish successful culture than those transported in Hibernate A. In vitro cultures were more successful from biopsies (84.2%) compared with autopsies (38.2%) and as monolayer on laminin-coated plates than as neurospheres. Primary cultures successfully established from autopsy samples were more likely to engraft in animal models than cultures established from biopsies (86.7% vs. 47.4%). Collectively, tumor engraftment was more successful when DIPG samples were directly implanted in mice (68%), rather than after culturing (40.7%). CONCLUSION: This multi-center study provides valuable information on the success rate of establishing patient-derived pre-clinical models of DIPG. The results can lead to further optimization of DIPG model development and ultimately assist in the investigation of new therapies for this aggressive pediatric brain tumor.

The current consensus on the clinical management of intracranial ependymoma and its distinct molecular variants.

Multiple independent genomic profiling efforts have recently identified clinically and molecularly distinct subgroups of ependymoma arising from all three anatomic compartments of the central nervous system (supratentorial brain, posterior fossa, and spinal cord). These advances motivated a consensus meeting to discuss: (1) the utility of current histologic grading criteria, (2) the integration of molecular-based stratification schemes in future clinical trials for patients with ependymoma and (3) current therapy in the context of molecular subgroups. Discussion at the meeting generated a series of consensus statements and recommendations from the attendees, which comment on the prognostic evaluation and treatment decisions of patients with intracranial ependymoma (WHO Grade II/III) based on the knowledge of its molecular subgroups. The major consensus among attendees was reached that treatment decisions for ependymoma (outside of clinical trials) should not be based on grading (II vs III). Supratentorial and posterior fossa ependymomas are distinct diseases, although the impact on therapy is still evolving. Molecular subgrouping should be part of all clinical trials henceforth.

Development of the SIOPE DIPG network, registry and imaging repository: a collaborative effort to optimize research into a rare and lethal disease.

Diffuse intrinsic pontine glioma (DIPG) is a rare and deadly childhood malignancy. After 40 years of mostly single-center, often non-randomized trials with variable patient inclusions, there has been no improvement in survival. It is therefore time for international collaboration in DIPG research, to provide new hope for children, parents and medical professionals fighting DIPG. In a first step towards collaboration, in 2011, a network of biologists and clinicians working in the field of DIPG was established within the European Society for Paediatric Oncology (SIOPE) Brain Tumour Group: the SIOPE DIPG Network. By bringing together biomedical professionals and parents as patient representatives, several collaborative DIPG-related projects have been realized. With help from experts in the fields of information technology, and legal advisors, an international, web-based comprehensive database was developed, The SIOPE DIPG Registry and Imaging Repository, to centrally collect data of DIPG patients. As for April 2016, clinical data as well as MR-scans of 694 patients have been entered into the SIOPE DIPG Registry/Imaging Repository. The median progression free survival is 6.0 months (95% Confidence Interval (CI) 5.6-6.4 months) and the median overall survival is 11.0 months (95% CI 10.5-11.5 months). At two and five years post-diagnosis, 10 and 2% of patients are alive, respectively. The establishment of the SIOPE DIPG Network and SIOPE DIPG Registry means a paradigm shift towards collaborative research into DIPG. This is seen as an essential first step towards understanding the disease, improving care and (ultimately) cure for children with DIPG.

A comprehensive evaluation of normalization methods for Illumina high-throughput RNA sequencing data analysis.

During the last 3 years, a number of approaches for the normalization of RNA sequencing data have emerged in the literature, differing both in the type of bias adjustment and in the statistical strategy adopted. However, as data continue to accumulate, there has been no clear consensus on the appropriate normalization method to be used or the impact of a chosen method on the downstream analysis. In this work, we focus on a comprehensive comparison of seven recently proposed normalization methods for the differential analysis of RNA-seq data, with an emphasis on the use of varied real and simulated datasets involving different species and experimental designs to represent data characteristics commonly observed in practice. Based on this comparison study, we propose practical recommendations on the appropriate normalization method to be used and its impact on the differential analysis of RNA-seq data.

Histone H3F3A and HIST1H3B K27M mutations define two subgroups of diffuse intrinsic pontine gliomas with different prognosis and phenotypes.

Diffuse intrinsic pontine glioma (DIPG) is the most severe paediatric solid tumour, with no significant therapeutic progress made in the past 50 years. Recent studies suggest that diffuse midline glioma, H3-K27M mutant, may comprise more than one biological entity. The aim of the study was to determine the clinical and biological variables that most impact their prognosis. Ninety-one patients with classically defined DIPG underwent a systematic stereotactic biopsy and were included in this observational retrospective study. Histone H3 genes mutations were assessed by immunochemistry and direct sequencing, whilst global gene expression profiling and chromosomal imbalances were determined by microarrays. A full description of the MRI findings at diagnosis and at relapse was integrated with the molecular profiling data and clinical outcome. All DIPG but one were found to harbour either a somatic H3-K27M mutation and/or loss of H3K27 trimethylation. We also discovered a novel K27M mutation in HIST2H3C, and a lysine-to-isoleucine substitution (K27I) in H3F3A, also creating a loss of trimethylation. Patients with tumours harbouring a K27M mutation in H3.3 (H3F3A) did not respond clinically to radiotherapy as well, relapsed significantly earlier and exhibited more metastatic recurrences than those in H3.1 (HIST1H3B/C). H3.3-K27M-mutated DIPG have a proneural/oligodendroglial phenotype and a pro-metastatic gene expression signature with PDGFRA activation, while H3.1-K27M-mutated tumours exhibit a mesenchymal/astrocytic phenotype and a pro-angiogenic/hypoxic signature supported by expression profiling and radiological findings. H3K27 alterations appear as the founding event in DIPG and the mutations in the two main histone H3 variants drive two distinct oncogenic programmes with potential specific therapeutic targets.

A Novel Sensory Wave (P25) in Myelin Oligodendrocyte Glycoprotein-induced Experimental Autoimmune Encephalomyelitis Murine Model.

Myelin oligodendrocyte glycoprotein (MOG)-induced experimental autoimmune encephalomyelitis (EAE) is a murine model for multiple sclerosis. This model is characterized by chronic and progressive demyelination, leading to impairment of motor function and paralysis. While the outcomes of the disease, including impaired motor function and immunological changes, are well-characterized, little is known about the impact of EAE on the electrophysiology of the motor and sensory systems. In this study, we assessed evoked potentials as a quantitative marker for in vivo monitoring of nervous system damage. Motor-evoked potentials (MEPs) and sensory-evoked potentials (SEPs) were first standardized in naïve C57BL mice and studied thoroughly in EAE mice. The duration of MEPs and the number of connotative potentials increased significantly alongside an increase in temporal SEP amplitudes. Moreover, a new SEP wave was identified in naïve animals, which significantly increased in MOG-induced EAE animals with no or mild symptoms (clinical score 0-2, 0-5 scale). This wave occurred ∼25 milliseconds poststimulation, thus named p25. P25 was correlated with increased vocalization and was also reduced in amplitude following treatment with morphine. As the EAE score progressed (clinical score 3-4, 0-5 scale), the amplitude of MEPs and SEPs decreased drastically. Our results demonstrate that desynchronized neural motor activity, along with hypersensitivity in the early stages of EAE, leads to a complete loss of motor and sensory functions in the late stages of the disease. The findings also suggest an increase in p25 amplitude before motor deficits appear, indicating SEP as a predictive marker for disease progression. PERSPECTIVE: This article assesses p25, a new sensory electrophysiology wave that correlates with pain-related behavior in MOG-induced EAE mice and appears prior to the clinical symptoms. Motor electrophysiology correlates with traditional motor behavior scoring and histology.

The pontine diffuse midline glioma, EGFR-subtype with ependymal features: Yet another face of diffuse midline glioma, H3K27-altered.

Diffuse midline glioma with two components: classical glial and ependymal.

"Hemispheric pilocytic astrocytoma" revisited: A comprehensive clinicopathological and molecular series emphasizing their overlap with other glioneuronal tumors.

Pilocytic astrocytomas (PA) typically exhibit distinct clinical, radiological, histopathological, and genetic features. DNA-methylation profiling distinguishes PA according to their location (infratentorial, midline, hemispheric, or spinal). In the hemispheric location, distinguishing PA from glioneuronal tumors remains a common diagnostic challenge for neuropathologists. Furthermore, the current version of the DKFZ classifier seems to have difficulty separating them from gangliogliomas. In this study, after central radiological review, we identified a histopathologically defined set of PA (histPA, n = 11) and a cohort of DNA-methylation defined PA (mcPA, n = 11). Nine out of the 11 histPA matched the methylation class of hemispheric PA, whereas 2 cases were classified at the end of the study as dysembryoplastic neuroepithelial tumors. Similarly, the mcPA cohort contained tumors mainly classified as PA (7/11), but 4 cases were classified as glioneuronal. The analysis of the 16 tumors with an integrated diagnosis of PA revealed that they affect mainly children with a wide spectrum of radiological, histopathological (i.e. a predominantly diffuse growth pattern), and genetic characteristics (large range of mitogen-activated protein kinase alterations). Based on these results, we consider hemispheric PA to be different from their counterparts in other locations and to overlap with other glioneuronal tumors, reinforcing the necessity of interpreting all data to obtain an accurate diagnosis.

Diffuse midline glioma invasion and metastasis rely on cell-autonomous signaling.

BACKGROUND: Diffuse midline gliomas (DMG) are pediatric tumors with negligible 2-year survival after diagnosis characterized by their ability to infiltrate the central nervous system. In the hope of controlling the local growth and slowing the disease, all patients receive radiotherapy. However, distant progression occurs frequently in DMG patients. Current clues as to what causes tumor infiltration circle mainly around the tumor microenvironment, but there are currently no known determinants to predict the degree of invasiveness. METHODS: In this study, we use patient-derived glioma stem cells (GSCs) to create patient-specific 3D avatars to model interindividual invasion and elucidate the cellular supporting mechanisms. RESULTS: We show that GSC models in 3D mirror the invasive behavior of the parental tumors, thus proving the ability of DMG to infiltrate as an autonomous characteristic of tumor cells. Furthermore, we distinguished 2 modes of migration, mesenchymal and ameboid-like, and associated the ameboid-like modality with GSCs derived from the most invasive tumors. Using transcriptomics of both organoids and primary tumors, we further characterized the invasive ameboid-like tumors as oligodendrocyte progenitor-like, with highly contractile cytoskeleton and reduced adhesion ability driven by crucial over-expression of bone morphogenetic pathway 7 (BMP7). Finally, we deciphered MEK, ERK, and Rho/ROCK kinases activated downstream of the BMP7 stimulation as actionable targets controlling tumor cell motility. CONCLUSIONS: Our findings identify 2 new therapeutic avenues. First, patient-derived GSCs represent a predictive tool for patient stratification in order to adapt irradiation strategies. Second, autocrine and short-range BMP7-related signaling becomes a druggable target to prevent DMG spread and metastasis.

Gliomatosis cerebri in children: A poor prognostic phenotype of diffuse gliomas with a distinct molecular profile.

BACKGROUND: The term Gliomatosis cerebri (GC), a radiology-defined highly infiltrating diffuse glioma, has been abandoned since molecular GC-associated features have not been established yet. METHODS: We conducted a multinational retrospective study of 104 children and adolescents with GC providing comprehensive clinical and (epi-)genetic characterization. RESULTS: Median overall survival (OS) was 15.5 months (interquartile range, 10.9-27.7) with a 2-years survival rate of 28%. Histopathological grading correlated significantly with median OS: CNS WHO grade II: 47.8 months (25.2-55.7); grade III: 15.9 months (11.4-26.3); grade IV: 10.4 months (8.8-14.4). By DNA methylation profiling (n=49), most tumors were classified as pediatric-type diffuse high-grade glioma (pedHGG), H3-/IDH-wildtype (n=31/49, 63.3%) with enriched subclasses pedHGG_RTK2 (n=19), pedHGG_A/B (n=6), and pedHGG_MYCN (n=5), but only one pedHGG_RTK1 case. Within the pedHGG, H3-/IDH-wildtype subgroup, recurrent alterations in EGFR (n=10) and BCOR (n=9) were identified. Additionally, we observed structural aberrations in chromosome 6 in 16/49 tumors (32.7%) across tumor types. In the pedHGG, H3-/IDH-wildtype subgroup TP53 alterations had a significant negative effect on OS. CONCLUSION: Contrary to previous studies, our representative pediatric GC study provides evidence that GC has a strong predilection to arise on the background of specific molecular features (especially pedHGG_RTK2, pedHGG_A/B, EGFR and BCOR mutations, chromosome 6 rearrangements).

Learning a Markov Logic network for supervised gene regulatory network inference.

BACKGROUND: Gene regulatory network inference remains a challenging problem in systems biology despite the numerous approaches that have been proposed. When substantial knowledge on a gene regulatory network is already available, supervised network inference is appropriate. Such a method builds a binary classifier able to assign a class (Regulation/No regulation) to an ordered pair of genes. Once learnt, the pairwise classifier can be used to predict new regulations. In this work, we explore the framework of Markov Logic Networks (MLN) that combine features of probabilistic graphical models with the expressivity of first-order logic rules. RESULTS: We propose to learn a Markov Logic network, e.g. a set of weighted rules that conclude on the predicate "regulates", starting from a known gene regulatory network involved in the switch proliferation/differentiation of keratinocyte cells, a set of experimental transcriptomic data and various descriptions of genes all encoded into first-order logic. As training data are unbalanced, we use asymmetric bagging to learn a set of MLNs. The prediction of a new regulation can then be obtained by averaging predictions of individual MLNs. As a side contribution, we propose three in silico tests to assess the performance of any pairwise classifier in various network inference tasks on real datasets. A first test consists of measuring the average performance on balanced edge prediction problem; a second one deals with the ability of the classifier, once enhanced by asymmetric bagging, to update a given network. Finally our main result concerns a third test that measures the ability of the method to predict regulations with a new set of genes. As expected, MLN, when provided with only numerical discretized gene expression data, does not perform as well as a pairwise SVM in terms of AUPR. However, when a more complete description of gene properties is provided by heterogeneous sources, MLN achieves the same performance as a black-box model such as a pairwise SVM while providing relevant insights on the predictions. CONCLUSIONS: The numerical studies show that MLN achieves very good predictive performance while opening the door to some interpretability of the decisions. Besides the ability to suggest new regulations, such an approach allows to cross-validate experimental data with existing knowledge.